Annual Review of Immunology Volume 20 2002

P1: FRK

February 1, 2002

12:47

Annual Reviews

AR152-FM

Annual Review of Immunology Volume 20, 2002

Annu. Rev. Immunol. 2002.20:1-28. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

CONTENTS FRONTISPIECE, Charles A. Janeway, Jr. A TRIP THROUGH MY LIFE WITH AN IMMUNOLOGICAL THEME, Charles A. Janeway, Jr.

xiv 1

THE B7 FAMILY OF LIGANDS AND ITS RECEPTORS: NEW PATHWAYS FOR COSTIMULATION AND INHIBITION OF IMMUNE RESPONSES, Beatriz M. Carreno and Mary Collins

29

MAP KINASES IN THE IMMUNE RESPONSE, Chen Dong, Roger J. Davis, and Richard A. Flavell

PROSPECTS FOR VACCINE PROTECTION AGAINST HIV-1 INFECTION AND AIDS, Norman L. Letvin, Dan H. Barouch, and David C. Montefiori T CELL RESPONSE IN EXPERIMENTAL AUTOIMMUNE ENCEPHALOMYELITIS (EAE): ROLE OF SELF AND CROSS-REACTIVE ANTIGENS IN SHAPING, TUNING, AND REGULATING THE AUTOPATHOGENIC T CELL REPERTOIRE, Vijay K. Kuchroo, Ana C. Anderson, Hanspeter Waldner, Markus Munder, Estelle Bettelli, and Lindsay B. Nicholson

55 73

101

NEUROENDOCRINE REGULATION OF IMMUNITY, Jeanette I. Webster, Leonardo Tonelli, and Esther M. Sternberg

125

MOLECULAR MECHANISM OF CLASS SWITCH RECOMBINATION: LINKAGE WITH SOMATIC HYPERMUTATION, Tasuku Honjo, Kazuo Kinoshita, and Masamichi Muramatsu

165

INNATE IMMUNE RECOGNITION, Charles A. Janeway, Jr. and Ruslan Medzhitov

KIR: DIVERSE, RAPIDLY EVOLVING RECEPTORS OF INNATE AND ADAPTIVE IMMUNITY, Carlos Vilches and Peter Parham ORIGINS AND FUNCTIONS OF B-1 CELLS WITH NOTES ON THE ROLE OF CD5, Robert Berland and Henry H. Wortis E PROTEIN FUNCTION IN LYMPHOCYTE DEVELOPMENT, Melanie W. Quong, William J. Romanow, and Cornelis Murre

197 217 253 301

LYMPHOCYTE-MEDIATED CYTOTOXICITY, John H. Russell and Timothy J. Ley

SIGNAL TRANSDUCTION MEDIATED BY THE T CELL ANTIGEN x

323

P1: FRK

February 1, 2002

12:47

Annual Reviews

AR152-FM

CONTENTS

xi

RECEPTOR: THE ROLE OF ADAPTER PROTEINS, Lawrence E. Samelson INTERACTION OF HEAT SHOCK PROTEINS WITH PEPTIDES AND ANTIGEN PRESENTING CELLS: CHAPERONING OF THE INNATE AND ADAPTIVE IMMUNE RESPONSES, Pramod Srivastava CHROMATIN STRUCTURE AND GENE REGULATION IN THE IMMUNE SYSTEM, Stephen T. Smale and Amanda G. Fisher PRODUCING NATURE’S GENE-CHIPS: THE GENERATION OF PEPTIDES FOR DISPLAY BY MHC CLASS I MOLECULES, Nilabh Shastri, Susan

371

Annu. Rev. Immunol. 2002.20:1-28. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

Schwab, and Thomas Serwold

395 427

463

THE IMMUNOLOGY OF MUCOSAL MODELS OF INFLAMMATION, Warren Strober, Ivan J. Fuss, and Richard S. Blumberg T CELL MEMORY, Jonathan Sprent and Charles D. Surh

GENETIC DISSECTION OF IMMUNITY TO MYCOBACTERIA: THE HUMAN MODEL, Jean-Laurent Casanova and Laurent Abel ANTIGEN PRESENTATION AND T CELL STIMULATION BY DENDRITIC CELLS, Pierre Guermonprez, Jenny Valladeau, Laurence Zitvogel, Clotilde Th´ery, and Sebastian Amigorena

495 551 581

621

NEGATIVE REGULATION OF IMMUNORECEPTOR SIGNALING, Andr´e Veillette, Sylvain Latour, and Dominique Davidson

669

CPG MOTIFS IN BACTERIAL DNA AND THEIR IMMUNE EFFECTS, Arthur M. Krieg

PROTEIN KINASE Cθ

709 IN

T CELL ACTIVATION, Noah Isakov and Amnon

Altman

RANK-L AND RANK: T CELLS, BONE LOSS, AND MAMMALIAN EVOLUTION, Lars E. Theill, William J. Boyle, and Josef M. Penninger PHAGOCYTOSIS OF MICROBES: COMPLEXITY IN ACTION, David M. Underhill and Adrian Ozinsky

761 795 825

STRUCTURE AND FUNCTION OF NATURAL KILLER CELL RECEPTORS: MULTIPLE MOLECULAR SOLUTIONS TO SELF, NONSELF DISCRIMINATION, Kannan Natarajan, Nazzareno Dimasi, Jian Wang, Roy A. Mariuzza, and David H. Margulies

853

INDEXES Subject Index Cumulative Index of Contributing Authors, Volumes 1–20 Cumulative Index of Chapter Titles, Volumes 1–20

ERRATA An online log of corrections to Annual Review of Immunology chapters (if any, 1997 to the present) may be found at http://immunol.annualreviews.org/

887 915 925

Annu. Rev. Immunol. 2002.20:1-28. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

P1: FRK

February 1, 2002 12:47 Annual Reviews AR152-FM

8 Feb 2002

9:28

AR

AR152-01.tex

AR152-01.SGM

LaTeX2e(2001/05/10) P1: GSR 10.1146/annurev.immunol.20.080801.102422

Annu. Rev. Immunol. 2002. 20:1–28 DOI: 10.1146/annurev.immunol.20.080801.102422 c 2002 by Annual Reviews. All rights reserved Copyright °

A TRIP THROUGH MY LIFE WITH AN IMMUNOLOGICAL THEME

Annu. Rev. Immunol. 2002.20:1-28. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

Charles A. Janeway, Jr. Section of Immunobiology, Yale University School of Medicine and Howard Hughes Medical Institute, 310 Cedar Street, New Haven, Connecticut 06520-8011; e-mail: [email protected]

KEYWDGRP> innate immunity, Toll-like reception, positive selection of T cells and B cells, suppressor T cells, adaptive immunity ■ Abstract In this essay, I make four points about the operation of the immune system. First, thanks to the innate immune system’s regulation of the main costimulatory molecules CD80 and CD86, the immune system rarely mistakes a pathogen for a self-antigen. Second, the adaptive immune system consisting of T lymphocytes and B lymphocytes can mistake self for non-self because adaptive immunity is selected in single somatic cells. Third, the adaptive immune system of T lymphocytes and B lymphocytes is always referential to self, as it is selected on self-ligands; it persists in the periphery on self-ligands; and at least for T cells, it is dependent on self-ligands to be able to mount a response. Fourth, it is becoming clear that regulatory or suppressor T cells are our main defense against autoimmunity, as my first boss, Richard Gershon, had predicted. These cells recognize antigen as do all T cells, but they secrete the immunoregulatory cytokines IL-10 and TGFβ.

INTRODUCTION My early life was one of health and wealth, although at the time I did not appreciate it. I grew up in Weston, MA, first on a highway on the south side of town (Massachusetts Route 30) and later on the north side of town, in a house that my parents had built for them by an architect friend after a careful search for sites within the same town. Throughout my early years, I was surrounded by love, kindness, and friends, especially after we moved to our new house on Concord Road in Weston. This neighborhood was very cohesive, held together by an unusual alliance between my father and the family next door, the Cummings family. They had bought the estate of the late Eddie Collins, a Hall of Fame second baseman for the Boston Red Sox, and over the ensuing years, they sold parcels to various people who became our neighbors, all about the same age and all with children. I made several very close friends, some of whom I still keep up with today, especially Kim Cummings. I had one other close friend named Rowley Elliston from my time on the south side of town who was then and continues to be my best friend. A medical experiment that I remember from my childhood, and still show 0732-0582/02/0407-0001$14.00

1

5 Feb 2002

12:48

Annu. Rev. Immunol. 2002.20:1-28. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

2

AR

AR152-01.tex

AR152-01.SGM

LaTeX2e(2001/05/10)

P1: GSR

JANEWAY

A medical experiment that I remember from my childhood, and still show to medical students now, happened when I contracted measles, which my father diagnosed by seeing Kopplick’s spots in my pharynx. At the time he was working on agammaglobulinemia and treating it by injection of intramuscular gamma globulin. He lined up all the kids in the neighborhood, injected them with gamma globulin, and then systematically exposed each of them to me. Each one developed protection against the measles virus but not measles itself. Imagine getting that past an IRB today! This made a strong impression on me, but I did not understand it at the time or I should have become an infectious disease specialist rather then an immunologist. But the impression stayed with me throughout my life and is still with me today. The lesson that I draw from this, and later learned in medical school, was that humoral immunity exists to protect the body from external threats to its integrity. This also illustrates how tightly knit a community we were on Concord Road. How privileged we were to be able to go out alone at night and play with one another. It is hard to imagine parents letting their kids run around in the dark nowadays! Of course, as the town had a reputation for good schools, I went to school at the Weston Public Schools, first to elementary school, then to Junior High School, which in those days was housed in the same building as the Public High School, and then I went on to Weston High School. Thus, it was a surprise when my mother wanted me to go to a private school because I was so happy and doing so well in public school. I still do not know why she wanted me out of the house. Perhaps it was the graduation of my friend Kim from Weston High School, or perhaps it was the departure of my best friend Rowley to the Cambridge School of Weston, but she insisted that I go interview at two private schools. One of these was the Phillips Exeter Academy, in Exeter, New Hampshire, and the other was Milton Academy, in Milton, Massachusetts. I remember interviewing at Exeter and telling my interviewer, who asked me why I wanted to attend the academy, that I was not at all sure that I did. Perhaps this unexpected answer may have convinced them that I was at least honest enough to admit to them that I was unsure. I don’t know if this is true, but in any case they accepted me. Thus began two years of what felt like prison to me, and to many of my classmates, as I learned recently at my 40th reunion at Exeter. This came as somewhat of a revelation to me, but I had seen it earlier, at my 20th reunion. At this reunion, all the guys who looked just as they had during their years at Exeter turned out to have been happy there, and all the guys who looked totally transformed had been miserable. It was the uniformity of unhappiness that came out at the 40th reunion that struck me and all of my classmates. There was not one of the happy camper types there, or else they were well hidden. Suffice it to say that I hated being at Exeter. This was before they admitted girls, which was part of the problem with the school. In any case, I managed to get a good education there and got advanced placement in three subjects, so I could go through Harvard College in three years and go on to my calling, which was in medicine, like my father before me, and his father and grandfather before him.

5 Feb 2002

12:48

AR

AR152-01.tex

AR152-01.SGM

LaTeX2e(2001/05/10)

P1: GSR

Annu. Rev. Immunol. 2002.20:1-28. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

A TRIP THROUGH MY LIFE

3

I was very lucky with my professors at Harvard College. In my freshman (or sophomore) year, I had two Nobel Prize winners, one of whom had already won the Prize, and one who was to win it later when I was living in Sweden. I then again got to marvel at his insights into the nature of chemical bonds. This was Eugene Lipscomb, and he taught those of us who were willing to listen to him all we needed to know about the nature of chemical bonds, using boron as his model molecule. It was for his work on boron that he won the 1976 Nobel Prize in chemistry, and he came to Sweden to give his lecture. I was then living in the city of Uppsala, Sweden, which is about an hour’s train trip north of Stockholm. It was wonderful to see him give the same flare to his Nobel Lecture that infused his lectures when I was a freshman chemistry student. After learning all about chemical bonds from Lipscomb, I found it relatively easy to breeze through the chemistry curriculum at Harvard, except for the labs, which still required some application. The second professor from whom I learned a great deal was George Wald, who taught freshman Biology, or Nat Sci 5 as it was called. He was particularly brilliant at explaining the central dogma of biology, which had just been formulated a few years before. This was that the genes were arranged on chromosomes and encoded information in DNA. The DNA code had to be transcribed into RNA, which was then translated into protein on ribosomes. This was really new at the time, and I was carried away with the excitement of these concepts. It is probably to George Wald that I owe my career in research, although there were many other central influences along the way: my lab instructor in Nat Sci 5, Hannah Gould, John Humphrey, Robin Coombs, Bill Paul, Dick Gershon, Susumu Tonegawa, and my many colleagues at Yale, to which I went in 1977. I want to give special credit to my wife, Kim Bottomly, to whom I would like to dedicate this reminiscence.

MY TRAINING AT MEDICAL SCHOOL AND IN IMMUNOLOGY When I finished Exeter and Harvard College, I knew that my destiny was to go to medical school, so I applied to several schools that had good reputations: Harvard, Johns Hopkins, and Yale. I never heard back from Yale, for some unknown reason, but I heard from Hopkins and from Harvard, which both accepted me. As I had a girl friend, who was later to become my wife, going to Wellesley College, that made the decision to go to Harvard Medical School easy. I worked hard there for the first two years, but I had trouble with the Biochemistry course, because I always solved problems by going back to first principles, whereas the instructors wanted me to use formulas that had been developed to make calculations easier for clinicians. But in any case, I basically learned my stuff in the first two years of medical school, and I passed with flying colors. I don’t actually know this, as we did not receive grades, but I could tell that I was doing well. The only thing that was wrong with this was

5 Feb 2002

12:48

Annu. Rev. Immunol. 2002.20:1-28. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

4

AR

AR152-01.tex

AR152-01.SGM

LaTeX2e(2001/05/10)

P1: GSR

JANEWAY

that I forgot how to think; I felt like a sausage casing that had been stuffed full every day. This image stayed in my mind for the rest of my life and became useful later when I started to teach medical students at Yale. I think it made me a better teacher than I would have been if I had not been exposed to professors who talked fast and convincingly about all manner of subjects. The only courses I remember enjoying, besides Immunology, were the courses in anatomy and neurobiology. Anatomy was my first encounter with the human body and with teams of students dissecting it. I used Gray’s Anatomy to study for this course. The reason I liked Gray’s Anatomy was that it was well written, which may have influenced my later decision to write my own book on immunology, now in its fifth edition. The course in neurobiology was taught by a number of young, enthusiastic professors under the guidance of Steven Kuffler. This course challenged one to think, and think we did. We would come up with experiments to do and new ideas, rather than the dreary memorization that characterized most courses. Another problem was that we had basically no free time, the days went on forever, and the nights were spent studying what we had heard during the day. In the middle of my second year at Harvard Medical School, I got a message from my lab instructor in Nat. Sci. 5, Hannah Gould, who asked me if I would like to work with her on globin synthesis, which she was studying in London. This piqued my interest because I had visited London in 1956 with my family and had fallen in love with the city and England in general. I visited it again with my friend Rowley Elliston in 1960. So I went to the Dean of Student Affairs, whose name was Joseph Gardella, and I asked him if he thought this was a good idea. He told me that I could go anywhere I wanted. That sent me back to the drawing board, and because I had been interested in immunology, I began seeking positions in immunology in London. My lab instructor in the immunology part of the course in microbiology was Hugh McDevitt, and he had had us read a series of papers on the origin of antibody diversity that had me really excited about this subject. So I mentioned that I would like to work in London in the same laboratory that Hugh had just returned from. I asked Hugh to arrange a position there for me, and after some hunting, he found that Brigitta (Ite) Askonas had planned to have a woman postdoctoral fellow who had, at the last minute, dropped out. I went to work with Ite, but when I arrived, it turned out that John Humphrey had a lab for me, so in the end I worked for John. It was a great two years as it turned out, and I wrote several papers with John and with Michael Sela, who had provided the synthetic polypeptides that I worked on. These were called L-TGA (for the left rotating amino acids tyrosine, glutamic acid, and alanine) and D-TGA (for the rightward rotating enantiomers of the same amino acids). This was before the invention of the single letter code, which would have listed them as L-YEA, a neater acronym than L-TGA. I compared the response to L-YEA with the response to D-YEA, which gave very low but not insignificant responses. These findings took me two years to compile and confirm, but in the end I wrote up four papers, all of which were accepted for publication.

5 Feb 2002

12:48

AR

AR152-01.tex

AR152-01.SGM

LaTeX2e(2001/05/10)

P1: GSR

Annu. Rev. Immunol. 2002.20:1-28. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

A TRIP THROUGH MY LIFE

5

Sally, my first wife and the mother of my daughter Katherine Anne, and I did other things while we were there. We traveled to Europe several times, to Wales and to Scotland, where by chance we met Chris Elson and his wife, talked about immunology and highland hikes, and ended up getting a ride back to Edinburgh. Overall, I had a good time. The only thing we could measure in those days was antibody production, and we measured it in great detail. It turned out that antibodies to L-YEA did not crossreact with D-YEA, and antibodies to D-YEA did not bind to L-YEA. That meant that they were specific. But how did they come by their specificity? Was a gene for each specificity found in the genome of the mice, or was there somatic mutation of a few or a single gene? John Humphrey himself was addressing the same question, in his case trying to rule out the “template hypothesis,” which proposed that the antigen was taken up by the lymphocytes and used as a template for the formation of specific antibodies. It was a wonderful environment for me to work in. After two years in John Humphrey’s lab, you can imagine my shock upon returning to medical school, where I discovered that what was called evidence in clinical medicine did not measure up to scientific standards. It was mainly guesswork and listening to one’s patients. I must say I was excited to see real live patients, and even some on whom I was allowed to perform simple procedures such as lumbar taps to measure the pressure in cerebrospinal fluid. I even performed a herniorraphy on a patient in the West Roxbury, VA, hospital where I did part of my surgical rotation. I had a really neat surgical instructor in my “Introduction to the Clinic” course at the beginning of my third year of medical school, so I thought for a while that I wanted to be like him. But after a while, I thought that going into surgery was going to condemn me to a life of routine procedures, so I looked elsewhere. I remember patients from those days quite clearly, especially one diabetic woman who went into a diabetic coma. She was on insulin treatment, and while Ralph Steinman (my former classmate and now fellow member of the National Academy of Sciences) bent over the woman, I asked for 50 ml of 50% glucose in water. The result of injecting this was very dramatic; the woman revived almost immediately, as Judah Folkmann had taught us she would, and Ralph was startled by the effect. But the main thing I remember was the endless hours of operating while functioning on adrenaline instead of glucose, due to the imperfect state of medical practice in those days, as well as in the present (though it has improved in many areas). The patients loved us for what we told them, not for what we could do for them. I began to feel as if my soul was rotting, and that was very distressing to me. I decided that I needed another dose of research, where no one would take my word for anything, and the results were always solid. Therefore, I began to ask around about research opportunities, especially when I realized that two of the elective courses I wanted to take in the second semester of my fourth year of medical school were not on offer that year. I had devised a trick to mark red blood cells, and I used it to show that blood group–specific antibodies could cross-link

5 Feb 2002

12:48

Annu. Rev. Immunol. 2002.20:1-28. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

6

AR

AR152-01.tex

AR152-01.SGM

LaTeX2e(2001/05/10)

P1: GSR

JANEWAY

A RBC but not B RBC when a mixture of A and B RBCs were agglutinated with anti-A antibody. Using this as a bargaining chip, I ended up in Robin Coombs’s laboratory on Tennis Court Road in Cambridge, England, on my fourth trip to this fabled isle. During my visit with Robin Coombs, several very significant events occurred. The first was that I made a friendship, which I still treasure, with a wonderful Scott of Pictish origin named Ian McConnell. He took me under his wing, and he hosted Sally and I to several nice dinners, prepared by his wife, Anna McConnell. Ian was irreverent in a way that would not be tolerated in an American lab, and this taught me a lesson that I have long remembered: Don’t be stuck on yourself. The second event was the arrival one day, in a swirl of fur and knee-length boots, of Phillippa, or Pippa, Marrack, one of the best known English/American immunologists. Her grandfather was J. R. R. Marrack, a noted immunologist who devised the technique known as equilibrium dialysis to measure the affinity of antibodies. At the first meeting, I was overwhelmed by her attractiveness, but I did not let my feelings out. However, over the years, we have had frequent encounters, most of them extremely positive. She has become the model of success that most female and male scientists aspire to. I consider her to be one of my close friends in immunology. While in Coombs’s lab, I worked on B cell receptors, using a rosetting technique that Robin had worked out with Phillip Gell. Coombs had made antiimmunoglobulin antibodies that were first used to show that the hemolytic disease of the newborn was due to antibodies to Rh blood group antigens. The basic idea was to add anti-immunoglobulin antibody to rabbit lymphocytes and antibodycoated sheep erythrocytes, and then mix them in a tube and look for the number of rosetted cells. We also took advantage of rabbit allotypes to confirm the specificity of our observations. This was the first demonstration that a subset of peripheral lymphocytes had immunoglobulin on their surface; we also used rabbit thymocytes to show that these cells did not form rosettes, which we interpreted to mean that there was no immunoglobulin on their surface. Similar studies using fluorescent antibodies were published shortly after we presented our results at the annual meeting of the British Society of Immunology. In comparing the submission dates of these two papers, it is clear that ours was submitted before the meeting and theirs after, but theirs was published first, while ours, although not revised, was published months later. I learned an important lesson from this: It is not he who submits first that gets priority, but he who gets published first.

INTERNSHIP AT THE PETER BENT BRIGHAM HOSPITAL When my days in England were over (I did not attend my own graduation), I came home and moved into our old apartment. The next day, I began my internship at the Peter Bent Brigham Hospital (PBBH, now Brigham and Women’s). I had already planned a postdoctoral fellowship in Baruj Bennacerraf’s laboratory at the NIH as one of many “yellow berets.” This term was coined to describe people who avoided

5 Feb 2002

12:48

AR

AR152-01.tex

AR152-01.SGM

LaTeX2e(2001/05/10)

P1: GSR

Annu. Rev. Immunol. 2002.20:1-28. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

A TRIP THROUGH MY LIFE

7

military service by going into the Public Health Service; I still have my honorable discharge certificate from the Public Health Service tucked away somewhere in my belongings. The year at the PBBH was at the same time one of the best and worst experiences of my life. What was good about it was serving my patients; what was bad was that I was forced to neglect my wife Sally to do so. In the end, we had a child and got divorced shortly thereafter, so I felt lonely for many years. However, there was a benefit to being single: I felt that I was free to do whatever I pleased. I had spent so much of my life living according to one or another person’s rules (the worst was at Exeter), and I needed time and freedom to rediscover myself and my own life. I know this will sound selfish to many who read this, but that was the way that I felt.

POSTDOCTORAL STUDIES UNDER WILLIAM E. PAUL AND HANS WIGZELL So for the next period of my life, I lived on my own and stayed up until all hours of the night. I was working hard, doing experiments that I believed in, getting results that I could interpret, and spending every other weekend with my young daughter Katherine Anne, whom I loved and still love very much. She was one year old when Sally and I split up. I was also living in an old log cabin on a dirt road outside of the town of Harmony, MD, that Sally and I had bought and fixed up. Eventually, when my postdoctoral years were over, I sold the property for about five times what it had cost. My daughter, Katie, could not understand why I would sell the house she knew as home. But I needed the money, which I later used to buy a home in New Haven, CT. During my postdoctoral fellowship at the Laboratory of Immunology, National Institute of Allergy and Immune Diseases, National Institutes of Health, I had the opportunity to work with several people who shaped my career, but the most important was Bill Paul, now head of the NIH Laboratory of Immunology, whom we called “the WEP” behind his back. This nickname was an affectionate abbreviation based on his initials. I first planned to work under Benacerraf, but he left the NIH the same day that I arrived from Harvard, leaving Bill in charge. Bill suggested that I work with Rose Lieberman, which I did for about a year, but it seemed that I was plowing old turf, and I wanted novelty. So Bill suggested that I try conjugating the hapten DNP to the heat-killed Mycobacterium tuberculosis known as H37Ra. I tried immunizing guinea pigs with this antigen, and I was surprised that I got what looked like hapten-specific T cell proliferation as a result of this immunization. I worked on this project for some time, and we published three or four papers on it over the five years that I spent happily in his lab. I also worked on a model of what was purported to be hapten-specific helper T cells, but it turned out to be an effect of antibody, as it could be passively transferred with antibody molecules purified over affinity columns.

5 Feb 2002

12:48

Annu. Rev. Immunol. 2002.20:1-28. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

8

AR

AR152-01.tex

AR152-01.SGM

LaTeX2e(2001/05/10)

P1: GSR

JANEWAY

I was feeling restless about my progress in science, and I remember having serious conversations with several senior scientists about what I should do with my life. Should I go back to medicine, which I loved because my patients seemed to appreciate my spending time with them? But in medicine I felt I was taking the easier pathway, the easy way out, whereas if I went on with research, I would always face skeptical colleagues who did not believe my interpretations of my data nor I of theirs. This felt much healthier to me, and so I decided to go into full-time research; I could always get further training in internal medicine if my plans did not work out. After five plus years at the National Institutes of Health, I was looking at offers from Harvard, again with Baruj Bennacerraf, Johns Hopkins with Roberto Poljak, Washington University in St. Louis with Herman Eisen, or at Yale with Dick Gershon. Of all of these offers, I chose Yale because the opportunities seemed more wide open; there were only a few immunologists at Yale, and I felt more comfortable with them than with any other group. I presented a seminar on my work with hapten-specific T cells, which we had already demonstrated to recognize both the hapten and various peptides, later confirmed by a German scientist (1). I was offered a job by Vincent Marchesi, Chairman of Pathology, at a starting salary of $27,500. Although all the other programs made better offers to me, I decided to go with my gut feelings, and I am glad that I did. But before I could come to Yale, Vince told me that he had to have a lab renovated for me, and for two other faculty who would join me; they were not selected at the time of my hiring. Before I began my career at Yale Medical School, I wanted further training, so I decided I would go to work with Hans Wigzell, then a Professor of Immunology in Uppsala, Sweden. I was asked by my former wife if I could take Katie with me, so Katie and I flew to England, and then on to Sweden in the fall of 1975. When we arrived, I was met at the airport by Hans Binz, who I had earlier met at a scientific meeting in the former country of East Germany. It was the first scientific meeting to be held in East Germany, and I remember crossing the border into East Germany, where the border police were most interested in finding newspapers, which they confiscated as western propaganda. I also remember having to translate for the East German Minister of Health on a drunken evening in the Rasthskellar of the hotel we were staying in, which was in a former hunting lodge of the Kaiser. This was a very interesting experience for me, as I had not spoken a word of German since 1963, when I graduated from college. My suspicion was that the health minister knew English but was too proud to let on, and that was how I was roped into this conversation. I then flew back to England, collected Katie from Liz Simpson, who had kindly taken us in, and flew to Sweden. My initial impression of Sweden was not good. First, I thought I was going blind, as the evenings got shorter and shorter, so that I went to and from the lab in the dark. Second, everyone looked the same, and they all answered my question of why were they working with the same expression: “Ja, m˚an musta har pengar.” I translated this to mean: “Yes, one has to have money.” I later had a Swedish postdoctoral fellow who confirmed my translation. If you

5 Feb 2002

12:48

AR

AR152-01.tex

AR152-01.SGM

LaTeX2e(2001/05/10)

P1: GSR

Annu. Rev. Immunol. 2002.20:1-28. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

A TRIP THROUGH MY LIFE

9

asked an American the same question, you would get a completely different type of answer, full of commitment to a cause or a career that was meaningful to them. I don’t yet know which I prefer, the Swedish or the American answer. At least the Swedes were honest enough to tell you that they were working for the pay. I also saw an article in the International Herald-Tribune that said that Sweden had the highest per-capita income in Europe, with the lowest standard deviation between the rich and the poor, and I believe that was true at the time. I loved working for Hans Wigzell. I even played squash with him (he always beat me), because I could get him to slow down enough after the match so that we could have a conversation. Another trick I learned was to drive him to the airport in my old SAAB which, although it had a hole in its floor, ran beautifully, far more smoothly than Hans’s own car. These were times when I could talk to him about science and about the experiments that I wanted to do. Although I published very little from this time, I learned how to isolate T cells and to grow them in vitro, so that later I could clone them, which seemed a wonderful thing to do once it became possible. Hans was always asking if I thought I would like it in “Jale,” which was, I am sure, a deliberate mispronunciation of “Yale.” I told him that I was sure that I would. Later, when he had become Rektor of the Karolinska Institut, he actually put his American friends in an old jail that had been converted to a hotel. I was finally able to tell him: “Ja, jag a¨ lska varend i F˚angelska” (Yes, I love being in jail).

MY CAREER AT YALE MEDICAL SCHOOL When I first came to Yale University School of Medicine in 1977, I was attracted by two things. One was the presence of Richard Gershon, whose early death at 57 in 1983 was one of the saddest moments of my life. The second reason was the opportunity that was given to me to build up an immunology program complete with a training grant, an immunology course, and the opportunity to interact with students at all levels. What has kept me here all these years was funding from the Howard Hughes Medical Institute. This came about as a result of a dispute between my Chairman, Vincent Marchesi, and the Chairman of Internal Medicine, Dr. Sam Thier. This dispute, which should have taught me a great deal about how the medical school operated, but did not, was over who should get the “big Hughes” and who should get the “little Hughes.” The terms were for the number of dollars attached to each: the “little Hughes” paid about $50,000 per year, while the “big Hughes” was supposed to pay all one’s expenses. Eventually, my chairman lost this argument, so the “little Hughes” came to him to support an investigator in one of the three disciplines which the Howard Hughes Medical Institute was then sponsoring: genetics, metabolism, and immunology. As I was the only junior faculty member in his department that fit this description, I became an HHMI investigator. Later,

5 Feb 2002

12:48

Annu. Rev. Immunol. 2002.20:1-28. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

10

AR

AR152-01.tex

AR152-01.SGM

LaTeX2e(2001/05/10)

P1: GSR

JANEWAY

HHMI did not approve of the Department of Internal Medicine’s candidate, so the “big Hughes” went to Richard Gershon, giving us both HHMI positions. Because we had both HHMI positions at Yale, we became a force in the politics that surround money in academia. I like to think we used this influence well, or at least for the benefit of the school and for immunology research at Yale. Later, the Hughes Tool Company was sold to General Motors and HHMI got a stricter board of trustees, but as I was already an established investigator, I was kept at that rank during the time that each new regime came in at HHMI. Also recruited were Dr. Donal Murphy and Dr. Robert Rosenstein, both directly out of postdoctoral positions. We also recruited Al Bothwell, the first true molecular biologist in the Division of Immunology in the Department of Pathology. Don was my best friend at Yale, and we shared an office, which held a secretary and small offices for each of us. Because his father had been posted to New Haven during World War II, he had actually been born in New Haven, which was quite novel for a faculty member at Yale. It was a very upsetting moment for me when I could not persuade the tenure allotment committee, which was controlled by Dean Rosenberg, to offer Don a slot so that he could try for tenure. This happened because I was never given the powers of a chairman but was always referred to by Dean Rosenberg as “the senior immunologist at Yale.” The next significant event in my life was the recruitment of my wife, Dr. Kim Bottomly, Ph.D., to the faculty at Yale. She became my inspiration in all things, my career, scientific research, and teaching. I would often ask her about how to teach a particular topic, and she always had clever thoughts. I should also mention another side of her. She was, and is, a wonderful mother to my three daughters: Katie, who was with me when I met Kim, and Hannah and Megan, who were born in New Haven after Kim’s arrival at Yale. This brought much joy and much fighting to our house on Livingston Street, where we still live. Once Kim settled into our new house, we began making plans to have a family. It was a hectic time in our lives. I was beginning to be known, and Kim was too, for her discovery of what we called T helper cells (TH) because they could induce antibody responses, and T inflammatory cells (TI) because they caused inflammation at sites of antigen injection. These were later to become famous as Th2 cells and Th1 cells, the nomenclature proposed by Tim Mossman on the basis of the cytokines these cells could secrete. I still prefer Kim’s nomenclature because it fits the functions of these cells more closely, but the world has overtaken it. Then, just as we were settling in to New Haven, Dick Gershon was diagnosed with a small cell lung cancer, throwing all our plans up in the air. Dick was a heavy smoker, and he told his wife that this type of cancer was not caused by smoking, which was nonsense. But she believed him, and so she sued Yale and the HHMI for an outrageous sum of money, further deepening tensions between Yale, HHMI, and myself. Eventually, the suit was thrown out by a judge, Dick’s wife and young daughter left town, and we began talking among ourselves about whether we would remain independent or whether we would be

5 Feb 2002

12:48

AR

AR152-01.tex

AR152-01.SGM

LaTeX2e(2001/05/10)

P1: GSR

Annu. Rev. Immunol. 2002.20:1-28. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

A TRIP THROUGH MY LIFE

11

swallowed up by the Pathology Department, of which we were then a division. When I was asked by Leon Rosenberg M.D., then the Dean of the Medical School, to meet him in his office, I thought he would offer me Dick’s job. Imagine my surprise when I found Vincent Marchesi, the Chairman of Pathology, sitting in his office with him! I thought this was going to be a chat with the Dean, but it turned into a diatribe by him against me and my behavior, and a set of demands that would have ended Immunology as an independent subject at Yale Medical School, and within the University as well. I was stunned. After this, Vin asked me to have a cup of coffee with him in the medical school cafeteria, where he laid it on even thicker than Lee Rosenberg had. I was further stunned. I felt as if I had just been put through a car wash without a car to protect me from all the blistering that I had undergone. I left Dean Rosenberg’s office saying that I would have to talk to my colleagues in the Division of Immunology. I did so, and they reacted with horror at the idea of our group becoming part of Pathology with no independence at all. When I went back to Lee Rosenberg, I told him that there was no enthusiasm for his idea about the immunology program, which at that time consisted of several investigators, as well as joint appointees Nancy Ruddle, Ph.D., Phillip Askenase, M.D., and Adrian Hayday, Ph.D. We all felt that we could not work with the Pathology Department any longer, and we wanted our own department. Leon’s reaction was that either he had chosen the wrong person, or that he had had the wrong idea. It is a tribute to his intellect, which is formidable, that he came to the latter conclusion. He was driven to this by an HHMI site visit shortly thereafter which consisted of people that he had to respect: Most of them were already in the National Academy of Sciences. This site visit told him he was off the mark in melding immunology into pathology, and to give him his due, he listened. I know because he called me the next day at my home (it must have been a Saturday), and he screamed at me for the better part of an hour, accusing me of setting him up with my friends. Nothing could be farther from the truth. I had carefully hidden my difficulties with Dean Rosenburg from my friends on the site visit team. They were stunned by his complaints that I was being resistant to his wishes. In any case, he calmed down and invited me to have lunch with him a few weeks later, at which he proposed the formation of a Section of Molecular Immunobiology, which I said was fine. He also said that he had appointed a close friend to head the search for a new chair whom he said was the best immunologist he knew. This turned out to be wide of the mark, but the basic idea seemed to me a lot better than the alternative, as long as the Immunobiology was an equal partner with the Molecular. We shook hands over that and ended our lunch. It later proved necessary to have lunch with him again, this time with all of my colleagues backing me up, to complain that his friend was not running a satisfactory search for a chair of our new section. He was, in fact, looking in the wrong places, and even scheduling seminars in rooms in which classes were taking place. In fact, we were able to point this out to Dean Rosenberg right after lunch, when a candidate for the job

5 Feb 2002

12:48

12

AR

AR152-01.tex

AR152-01.SGM

LaTeX2e(2001/05/10)

P1: GSR

JANEWAY

was found waiting outside a full lecture hall! Eventually, we got our way when Leon hired Richard Flavell as Chair of a new Section of Immunobiology.

Annu. Rev. Immunol. 2002.20:1-28. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

MY TEACHING CAREER AT YALE During the period from 1977, when I arrived at Yale University School of Medicine, until around 1980, we taught immunology as a subdiscipline of pathology, with a team of teachers coming from several departments, including Bob Cone from Surgery, Nancy Ruddle from Epidemiology and Public Health, Phil Askenase and John Dwyer from Internal Medicine, and Dick, Don Murphy, Bob Rosenstein, and myself from the Department of Pathology. The teaching of the subject was very disjointed and lacked focus, so I volunteered to give all the lectures in a short course in the first year of Medical School. This worked very well, and soon the students were making a real effort to get to my classroom and especially to the tutorials that I designed to go with the lectures. Although I taught undergraduates, graduate students, and medical students, my favorite students were always the medical students. At first, this was not true, and I was harshly criticized by them in their course critiques, which all Yale medical students have to write. I eventually taught Immunology to Yale medical students for about 15 years. This was a very happy period in my life, as I had total control over the course material, the tutorials that went with them, and the composition of the exam. It was always a short answer exam, with two questions on a page, and six pages of questions, which I would then have my secretary divide up and give one page to each of my five tutorial leaders and one page to me. In this way, I could guarantee that there was no bias in the grading of the exam. I used to have a curve set up that told me which student was learning and which one was not. In the early years, I learned that those who attended the tutorials would always pass the exam, while those that did not would often fail the exam. Later, under Dean Rosenberg, the exams were made mandatory in order to force the students to attend lectures. I thought this was a bad move, and I spoke up at the faculty meeting where this was decided. I think I offended many of my colleagues who had to teach longer courses covering more material than I did, but I think I was right in stating that if students were not attending lectures, one had only one’s self to blame. Make your lectures interesting and lively and the students come in droves. That was my experience, and I eventually won a Bohmfalk Teaching Award for my efforts. This took time and effort on my part and that of my colleagues who taught the tutorials, but in the end, at least for me, it was worth the effort. The prize for the most imaginative use of the tutorials went to my wife, Kim, who nearly won her own Bohmfalk Award for her outstanding tutorial teaching. The students used to crowd into her room, so that over time I had to assign the biggest room in our teaching building to Kim. It was a period of great happiness for me. After a relatively short time, I was a very successful teacher. It was the receipt, in 1983, of the best text that I have yet read, The Molecular Biology of the Cell

5 Feb 2002

12:48

AR

AR152-01.tex

AR152-01.SGM

LaTeX2e(2001/05/10)

P1: GSR

Annu. Rev. Immunol. 2002.20:1-28. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

A TRIP THROUGH MY LIFE

13

(MBOC), that got me thinking about writing an immunology textbook. All the new information that was pouring out of my colleagues’ labs made me even more enthusiastic about writing my own book. By 1987, I sent a copy of my course notes to Gavin Borden at Garland Press, the publishers of MBOC. It was widely heralded as the best text ever published in cell biology. The reasons this book was so successful were two: First, a unique feature of the book was that it was written by a bunch of guys who loved their science and were unafraid of criticism. They sent all of their drafted chapters to world-renowned experts for a read-through, then modified them according to the comments they received. Second, this group was mother-henned by Miranda Robertson, the biology editor of Nature, at that time a leading journal in many areas of biology, including immunology. These two facts, along with Gavin’s good natured hosting of the authors in his house in a corner of London just off Abbey Road, made MBOC the most popular of cell biology texts. I picked up the essence of this book and adapted it to my immunology notes; shortly thereafter, I sent them to Gavin and thus began a long collaboration. Basically, our goal was to put Ivan Roitt out of the immunology textbook business, as up until that time, he had written the best book about the subject I loved so much. It turned out that Gavin knew Vitek Tracz, the originator of the Ivan Roitt publishing empire, and he set us up to work with Vitek. Another key player was soon on hand in the person of Miranda Robertson. So we had the whole MBOC team assembled to write a new textbook of immunology, except that we needed a person doing immunology in England; Miranda miraculously and brilliantly chose Paul Travers. Thus began a long collaboration between Paul, Miranda, Vitek, and myself, working out of Vitek’s London office. At first, I was happy to stay in the same house that the MBOC authors used, but it meant trooping across town on the London Underground, which was not ideal, and it also meant a series of long absences from home. These were mainly due to long wrangles with Miranda, usually over trivial results like complement lysis of red blood cells, which she thought was important and we did not. In any case, when the book was finally published in 1994, it was almost immediately adopted by many course directors, enabling me to relax about earning my keep in other ways. I was just trying to put Immunology in terms that I could understand, and it turned out that other people saw things much the way I did. Maybe I am just simpleminded, but, as it seemed to make sense to many brilliant students that I taught, it is likely to be more than that.

HOW MY THINKING ABOUT THE IMMUNE SYSTEM HAS CHANGED DURING MY YEARS AT YALE The main message I would like to convey is how I have grown intellectually during my years at Yale. As I said before, I was happy injecting antigens in complete Freund’s adjuvant or in alum with Bordetella pertussis, which gave me antibody production and later T cell proliferative responses. But I began to

5 Feb 2002

12:48

Annu. Rev. Immunol. 2002.20:1-28. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

14

AR

AR152-01.tex

AR152-01.SGM

LaTeX2e(2001/05/10)

P1: GSR

JANEWAY

think that there must be more to immunology than just these responses, and I wondered where these influences were coming from. Not all features of adaptive immunity could be accounted for by the adaptive immune system consisting of T lymphocytes and B lymphocytes; there must be something else contributing to the signals that were necessary for immune induction. Wherever one looked, there were signs of this: T-independent antibody formation, T-independent-macrophage activation, and complement activation, especially by the so-called alternative pathway of complement activation. This pathway was very active against microbes, but it was quiescent with the cells of the host. I thought this might be due to innate immunity, about which I knew very little, and later that idea turned out to be right. Most people at that time were studying adaptive immunity. Even people who were studying invertebrates, which we now know cannot mount an adaptive immune response because they lack the genes required to make immunoglobulins and T cell receptors, were trying to prove that they made vertebrate-like adaptive immune responses. This was later shown to be wrong, most importantly by my colleague David Schatz, Ph.D., who along with Marjory Ottinger, Ph.D., defined the RAG genes. He later published what I think of as one of the best papers I have read, in which he described the acquisition of the two closely linked RAG genes and the sites upon which they act, the recombination signal sequences (RSS) (2). These turned out to come from a retroposon that invaded the germ line of some lowly vertebrate, as only vertebrates have adaptive immune systems. I return to this later when I introduce adaptive immunity in mice and humans.

INNATE IMMUNE RECOGNITION OF PATHOGENS AS A FIRST STEP IN ADAPTIVE IMMUNITY We performed several experiments that I like to think are crucial to the development of the idea that adaptive immunity was dependent upon a functioning innate immune system. This led eventually to the cloning of the first pattern recognition receptor, hToll or, more appropriately, hTLR4, for human toll-like receptor-4. The first of these experiments was performed by Yang Liu, who now has an endowed chair as Professor of Pathology at Ohio State University, who demonstrated that the ligand for the T cell receptor (TCR), which in our case was anti-CD3, had to be presented by the same cell as the so-called costimulatory molecules (3). This study had to be published in the Proceedings of the National Academy of Sciences, as it showed only a “quantitative” difference, according to a review we received from Nature. This was true, but the quantitative difference was of the order of 100–1000-fold, so overcoming it either by increasing the level of the costimulatory molecules or of the TCR ligand could not be expected to happen in vivo. The reason for this is that the TCR ligand is made up of a complex of a selfMHC molecule with an antigenic peptide, and there are limited numbers of MHC molecules on a cell. It also requires at least 100–200 specific ligands to activate

5 Feb 2002

12:48

AR

AR152-01.tex

AR152-01.SGM

LaTeX2e(2001/05/10)

P1: GSR

Annu. Rev. Immunol. 2002.20:1-28. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

A TRIP THROUGH MY LIFE

15

a T cell. As the difference we found was in the neighborhood of 100–200-fold greater specific ligand when the ligand and the costimulator were presented on two different cells, it would be impossible to achieve this level. The finding that ligand and costimulator had to be presented by the same cell was subsequently confirmed by several other investigators using genuine T cell receptor (TCR) ligands, namely peptide:MHC complexes (4). What was not known at that time was what controlled the expression of the costimulatory molecules on the surface of an antigen-presenting cell. We thought it must be an infectious agent or a product of an infectious agent, because Yang and I had earlier shown that a whole variety of microbial substances could stimulate the antigen-presenting cells to express what we believed to be costimulatory activity (5). But I was not sure, so I pushed on with the theory that some innate immune receptor had to be responsible for the regulation of the costimulatory molecules expressed by adjuvant-primed antigenpresenting cells. This led to a very frustrating part of my life, in which a lot of work on adaptive immunity was done, but relatively little on innate immunity. Then, in a wonderful stroke of luck, I began receiving e-mails from a person by the name of Ruslan Medzhitov, who came from Tashkent in Uzbekistan and had spent an entire month’s stipend to obtain a copy of my introductory essay in the 1989 Cold Spring Harbor Symposium (6). I was interested in his ideas, but not overwhelmed with enthusiasm for hiring him. But then, in another stroke of blind luck, Ruslan earned a visiting fellowship to the University of California in San Diego, where he was to analyze protein sequences with Dr. Russell Doolittle. He went there and a short time later, I received a call from Dick Dutton, who was at that time the head of the biology department at UCSD and the president of the American Association of Immunologists. He had just met Ruslan, and he told me two things that convinced me to accept him in my laboratory. The first was that Ruslan still wanted to come to my lab. The second was that he was “a genius.” I immediately e-mailed Ruslan and told him that he could come to my lab. When Ruslan first arrived, he spent some time on various projects related to adaptive immunity. However, it was when he began turning his attentions to innate immunity that I knew that I had found the right colleague to test my ideas. He was what I had been waiting for all these years since I wrote the article in the Cold Spring Harbor Symposium: a man who was fixated on finding out what controlled the expression of costimulatory molecules on the surface of antigen-presenting cells. And even better than that, he was convinced that appropriate cytokines would be released as well. Ruslan had a degree in theoretical biochemistry, for which he had worked very hard, but he had no practical laboratory experience. I told him that he would need to develop such skills to be successful. I passed him one day in the lab, and I asked him how he was making out with the practical side of laboratory work. He said: “Fine, now I can tell a bottle of culture medium from a centrifuge.” I knew then that I had a winner, and I was right. Although I had trained many excellent graduate students and postdoctoral fellows, Ruslan was the first genuine superstar to work in my laboratory, and I have enjoyed interacting with him ever

5 Feb 2002

12:48

Annu. Rev. Immunol. 2002.20:1-28. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

16

AR

AR152-01.tex

AR152-01.SGM

LaTeX2e(2001/05/10)

P1: GSR

JANEWAY

since. He was hired as an assistant professor in the Section of Immunobiology in 1999, and he was subsequently made a Howard Hughes Medical Institute Assistant Investigator. I had never had such a postdoctoral fellow, and I don’t expect to have another. Ruslan and I published several papers over the years that he was in my lab, and afterward when he had his own lab, but the first was the most influential (7). In it, he described the cloning of a human homologue of the Drosophila Toll protein, which clearly induced B7.1 and B7.2 (or CD80 and CD86 as they are now called). This protein was later renamed hTLR4 for human toll-like receptor-4. It turns out there are 10 TLRs in the mouse and human genomes. The genes that encode each of these are more closely related to one another in the mouse and humans than they are to other TLRs within a species, so that one can move from mouse to human and back with ease. We knew from Jules Hoffman, with whom I have collaborated for several years, that defects in the Toll system at any level would render the fruit fly Drosophila melanogaster susceptible to overwhelming fungal infection and subsequent death. This led to a dramatic cover of the journal Cell, in which a fruit fly that was clearly dead had fungal hyphi coming out of its cuticle, due to a defect in the Toll signaling pathway (8). It was seeing this picture that led us to focus on Toll cloning, although Ruslan had already started to clone it. I urged him on to do this, and he outdid my exhortations. He published one of the most important papers in biology, as shown by its inclusion in Ben Lewin’s compendium of “Great Discoveries in Biology.” What this paper did was to answer a crucial question: How does the body know when to respond to a foreign antigen, and how does it know how to avoid responses directed at self-antigens? The mechanism that Ruslan discovered is a simple one: It does so by recognizing molecular patterns associated with the outer coats of bacteria. In the case of TLR4, the pathogen associated molecular pattern, or PAMP, is LPS. Other pathogens are recognized by other TLRs, such as TLR2 (9), which recognizes lipoteichoic acid found commonly in the coat of gram positive bacteria, TLR5, which recognizes flagellin (10), or the unmethylated CpG DNA found in bacteria but not in mammals by means of TLR9 (11).

THE ANALYSIS OF ADAPTIVE IMMUNITY IN RECENT TIMES Once I realized that the problem of self:non-self discrimination had been solved in principle by Ruslan’s magnificent discovery, I felt liberated to go back to my first love, which was the study of adaptive immunity. I knew from my childhood experiences with my father, the original discoverer of X-linked agammaglobulinemia (12), that the adaptive immune system was essential to living a healthy life. The discovery, in my lifetime, of several vaccines that could protect one from infection and could thus improve health, all depended on inducing an adaptive immune

5 Feb 2002

12:48

AR

AR152-01.tex

AR152-01.SGM

LaTeX2e(2001/05/10)

P1: GSR

Annu. Rev. Immunol. 2002.20:1-28. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

A TRIP THROUGH MY LIFE

17

response. Thus, I was left with the idea that both the innate immune system and the adaptive immune system were essential partners in living a long and healthy life (apart from the occasional cold). Of the two systems, I would rate the innate immune system the more important, as it is as old as the emergence of multicellular organisms from the “primordial slime,” and as defects in innate immunity are as rare as an undercooked steak. The requirement for the adaptive immune system of T cells and B cells was reinforced recently by the emergence of the acquired immune deficiency syndrome (AIDS), which attacks the adaptive immune system. It may also attack the innate immune system, as the initial infection occurs in dendritic cells or in macrophages (13). The rising toll of this deadly disease has given the lie to the easy assumption that we had beaten infectious diseases with the introduction of antibiotics, especially of the wonder drug penicillin. The discovery of antibiotics near the end of World War II led to the abolition of the Department of Microbiology at Yale and at several other highly regarded medical schools. But later, the Section of Molecular Pathogenesis was refounded in order to deal with epidemics like that of AIDS, antibiotic-resistant tuberculosis, and other infectious diseases. Unless we can devise a vaccine against HIV, the causative agent of AIDS, we will all become infected and eventually die. I used to show a cover of Newsweek which posed the question: “AIDS: The Public Health Threat of the Century?” And I would go on to say that I thought the public health threat of the century was living in the White House (Ronald Reagan was then the president). No longer do I find that joke to be funny; with well over 50,000,000 cases of HIV-1 infection worldwide, it is a true challenge that we must face or become extinct, like the Dodo bird. What AIDS does is to attack the adaptive immune system. Eventually, when the CD4 T cell count, normally around 5000 cells per milliliter, falls to 500 cells per milliliter or fewer, one dies of an infection with nonpathogenic bacteria, yeast, fungi such as Candida albicans, or from lymphomas that seem to be recognized and controlled by the adaptive immune system. We have learned a lot about the agent that causes this infection, but no one seems to have any idea how to make a vaccine to combat it. And time is running out. Thus, one needs both an innate immune system to watch out for pathogenic bacteria and viruses, and an adaptive immune system to make an effective response to those pathogens that can overcome the innate immune system. The other reason that we need an adaptive immune system is to protect us from re-infection with the same organism; such a system can make antibodies and T cells that can jump-start the adaptive immune response when the organism re-encounters the same or a closely related pathogen. This is a job that the innate immune system cannot do, as it has been selected over evolutionary time rather than in individual cells, as is true of the adaptive immune system. This means that the receptors for innate immunity are found in all multicellular organisms, whereas adaptive immunity is in evolutionary terms a new invention that is found uniquely in vertebrates.

5 Feb 2002

12:48

Annu. Rev. Immunol. 2002.20:1-28. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

18

AR

AR152-01.tex

AR152-01.SGM

LaTeX2e(2001/05/10)

P1: GSR

JANEWAY

So I began to look more carefully at the adaptive immune system, and I eventually came to the conclusion that I had it all wrong. The adaptive immune system of lymphocytes depends on self-antigens for its development in the thymus, for its longevity in the peripheral lymphoid tissues, and even for its activation. I have only recently come to this realization, so convinced was I that Burnet (14), who first proposed the clonal selection hypothesis, was correct. This was a revolutionary thought for me, and it is still working its way through my thick skull. But I am convinced that it is true, and we have much evidence that supports this idea. I would like to start with the T lymphocytes with which I am more familiar, but I will end up with B lymphocytes and with regulatory T cells, which are the subject that drew me to come to Yale in the first place. Regulatory T cells are just a neologism for suppressor T cells, which Richard Gershon originally, and I now think correctly, predicted were a unique lineage of T lymphocytes that were needed to protect the host from autoimmune attack. So I will call them suppressor T cells in honor of Richard Gershon, my first real boss (15).

THE POSITIVE SELECTION OF T LYMPHOCYTES ON SELF-PEPTIDE:SELF-MHC COMPLEXES T cells are generated in the thymus by a series of rearrangements of their receptor gene segments. Once the complete T cell receptor is put on the T cell surface, the T cell has to undergo two processes before it can emerge into the peripheral lymphoid tissues. These are conventionally called positive (16) and negative (17) selection. It is still not clear to me that we know the sequence of these reactions, but it is easier to consider positive selection as coming first, and negative selection as happening at various times during T cell development. Positive selection raises the level of the TCR significantly, thereby making the developing thymocyte more sensitive to stimulation by self-peptides. The really key point is that the only peptides available to the developing T cells are self-peptides, and so from the start T cells are positively (and negatively) selected on self-peptide:self-MHC complexes. This seems to be an inescapable conclusion to me, so I state it boldly and without exception. I may be wrong, but everything I think I have learned over the years points in this direction. If this is so, how do any T cells ever escape from the thymus? The T cells that escape from the thymus have highly variable receptors, and some of them only recognize the self-peptide:self-MHC complexes weakly, enough to drive positive selection but not enough to activate the thymocyte, which leads to apoptosis or activation-induced cell death. The fortunate T cells that have such receptors then emigrate from the thymus and fill up the periphery. At least 85% of the T cells fail to recognize any ligand at all, probably due to a combination of T cell receptor (TCR) variability and MHC polymorphism, and thus these developing T cells fail positive selection and die within the thymic cortex. A few thymocytes have such avid receptors that the thymocytes are activated upon encounter with the

5 Feb 2002

12:48

AR

AR152-01.tex

AR152-01.SGM

LaTeX2e(2001/05/10)

P1: GSR

Annu. Rev. Immunol. 2002.20:1-28. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

A TRIP THROUGH MY LIFE

19

self-peptide:self-MHC complex leading to activation-induced cell death. So the yield of thymocytes is between 1% and 5% of the initial numbers; for every thymocyte that is exported to the periphery, 20–100 thymocytes die. The upshot of these two processes is the selection of a TCR repertoire that is both self-MHC restricted and self-tolerant, as I have taught my students for years. So why do I think I see things differently now than I did when I arrived at Yale in 1977, 25 years ago? I now think that the recognition of self-ligands is important not only for selection of such a repertoire, but also for sustaining this repertoire in the periphery, and for signaling for activation when a pathogen or other antigen appears in the periphery. The reason I say a pathogen should be obvious to the informed reader, but I will reiterate that the presence of a pathogen or its products is essential to arm the antigen-presenting cells with the needed costimulatory molecules, as well as to stimulate the uptake, processing, and presentation of antigens. So that is how I now view the activation of T cells.

THE INTERACTION WITH SELF-PEPTIDE:SELF-MHC LIGANDS DRIVES PERIPHERAL T CELL SURVIVAL Peripheral T cells survive by interacting with a ligand that is almost certainly similar or identical to that which drove them to be positively selected in the thymus. The evidence for this is very strong, although some doubt it and claim that the homeostatic cell division observed in the periphery depends on the availability of so-called “space.” This impression is created by the use of either RAG−/− mice or irradiated recipients as hosts for the T cells. In our earlier experiments, we also used irradiated recipients to show that CD4 T cells could cycle to fill the space created by irradiation. However, we have now also performed identical experiments that show cycling of T cells in mice that are normal and not irradiated. Although the cycling was somewhat reduced, it still occurred. Thus, we believe that the T cells are recognizing self-peptide:self-MHC complexes on the surface of host cells (18; and C. Viret and C. A. Janeway, Jr., J. Immunol., in press). The cells involved in peripheral T cell survival are almost certainly the same cells that present foreign antigen to the T cell, that is dendritic cells (DCs). This was shown earlier by Brocker et al. (19), who expressed MHC class II molecules on MHC class II −/− dendritic cells under the control of a DC-specific promoter. These DCs are not activated by pathogens and therefore lack the costimulatory molecules or the foreign antigens necessary to trigger an adaptive immune response. Thus, these DCs can be considered as necessary for T cell survival but not for T cell activation by the self-peptide:self-MHC complexes they bear on their surfaces. I envision this being true for all T cells, as illustrated by the studies in Benedetta Rocha’s laboratory (20) for CD8 T cells, on which our original studies were modeled. We went on to examine the role of the peptide component and the specificity for MHC in our studies, also in mice that had been exposed to ionizing radiation (21).

5 Feb 2002

12:48

Annu. Rev. Immunol. 2002.20:1-28. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

20

AR

AR152-01.tex

AR152-01.SGM

LaTeX2e(2001/05/10)

P1: GSR

JANEWAY

We found that using allogeneic MHC molecules would not support survival, nor would deletion of the α chain of the peptide exchange factor H-2M mice support survival of MHC class II–restricted CD4 T cells. Thus, I am reasonably certain that it is self-peptide:self-MHC complexes that provide the low-grade signaling that we see in our TCR transgenic CD4 T cells. Other interpretations of CD8 cell survival have been made by other authors (21), but as we have seen similar results in unmanipulated recipients of CD4 T cells, I feel certain that these T cells, at least, survive by constantly making and breaking contact with DCs bearing the essential self-peptide:self-MHC ligands (18). Thus, CD4 T cells, which I have spent much of my life studying, are not only positively selected on self-peptide:self-MHC ligands, they persist on self-peptide:self-MHC ligands. The real open question now is whether the recognition of self-peptide:self-MHC ligands also contributes to T cell activation? We examine that question in the next segment of this essay.

THE ADAPTIVE IMMUNE RESPONSE TO FOREIGN ANTIGEN ALSO SHOWS SIGNS OF REQUIRING CONTACT WITH SELF-PEPTIDE:SELF-MHC LIGANDS When na¨ıve CD4 T cells are stimulated with agonist ligands, which consist of a specific foreign peptide bound to a particular self-MHC molecule, they conventionally give a full response. This includes, but is not limited to, proliferation, cytokine secretion, and differentiation into either Th1 CD4 T cells or Th2 CD4 T cells. These cells secrete different sets of cytokines: Th1 cells secrete interferon γ and TNFβ, whereas Th2 cells secrete IL-4, IL-5, and IL-13. These cells also share at least two cytokines in common, GM-CSF and IL-3. Each cytokine is recognized by a different receptor, and these patterns are used to type CD4 T cells into these two lineages. This is in response to an agonist ligand, which, during intrathymic development, can delete cognate T cells totally. So what happens if you present an agonist peptide in a pure form by attaching it to the β chain of the I-Ab molecule and putting the transgene into a mouse that lacks endogenous Aβb? This question was initially tested with wild-type TCRs. In order to guarantee that the peptide, which my laboratory initially identified as a dominant binding peptide of the Eα molecule, residues 52-68, is the only peptide available to the cells, these mice had to be on a background that lacked the MHC class II invariant chain. Under these conditions, one could demonstrate that all of the MHC class II molecules were modified by a single peptide, Eα52-68. This was demonstrated by showing that the staining of the cells in the spleen was as intense with the Y-Ae antibody that we had produced to this peptide bound to I-Ab, but these same cells also bound an equivalent number of molecules of another antibody, which in our case was called Y-3JP, specific for related epitope that bound to all I-Ab molecules. However, really convincing evidence for this conclusion involved the blocking of Y-Ae antibody by the Y-3JP antibody and vice-versa, which proves that such mice bear exclusively the I-Eα chain peptide bound to I-Ab.

5 Feb 2002

12:48

AR

AR152-01.tex

AR152-01.SGM

LaTeX2e(2001/05/10)

P1: GSR

Annu. Rev. Immunol. 2002.20:1-28. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

A TRIP THROUGH MY LIFE

21

The results of several studies by Kappler, Marrack, and Ignatowicz tell one several things. First, one derives a repertoire of T cell receptors that is basically diverse, although it is tolerant to the peptide:MHC complex carried artificially by this mouse (21). Second, there is a clear impact on the T cell receptor repertoire, in that many if not all of the T cells assayed as T cell hybrids have alloreactive potential. The commonest alloreactivity was to the I-Ab molecule itself, presumably by binding a variety of self-peptides and accounting for the majority of the T cell hybrids produced from these mice. The next commonest was reactivity to I-Ad, which has a very similar peptide binding groove that also binds a fair number of the same Eα peptides, although this could only be determined by mass-spectrometry analysis (22). These studies suggest that these mice inhibit negative selection to a host of self-peptide:self-MHC ligands, leading to reactivity to such ligands that normally select against such TCRs in the thymus and in the periphery. As we had studied a TCR that was strongly stimulated by the Eα52-68:I-Ab complex, we wondered whether it would be stimulated by this agonist peptide. Before I speak of the role of self-peptide:self-MHC complexes in T cell recognition, I need to discuss an experiment by Alam, Travers, and Gascoigne that appeared in Immunity in 1999 (23). In this study, the authors showed that agonist and antagonist peptide bound nearly identically to the several TCRs that they tested at room temperature (25◦ C), such that it seemed that differences in TCR binding could not explain the differences observed in functional assays, including thymic organ culture and peripheral T cell stimulation. However, when they ran their assays at physiological temperature (37◦ C), they got quite different results. They observed what they interpreted as TCR dimerization when they used an agonist peptide. These dimeric complexes were very stable, with a t1/2 of about 600 s, compared to a t1/2 of 20–30 s observed at 25◦ C. However, antagonist peptides, which can drive positive selection but also may inhibit responses to agonist peptides, had a t1/2 of 20–30 s whether they were assayed at room temperature or at the physiological temperature of 37◦ C. Thus, this assay clearly discriminated agonist from antagonist peptides. According to Alam et al. (23), their data fit best to a model of dimerization of the TCRs. But what would the second TCR recognize on a cell surface, where stimulation can occur with as few as 100–200 foreign peptides per cell? What seems most likely to me is that the second TCR would bind to a self-peptide:self-MHC complex. Such interactions have been known for many years in the response to self-APCs by autologous T cells in the so-called autologous MLR. This required the use of dendritic cells as stimulators for optimal responsiveness. The use of this assay has gone out of fashion, but it was studied by Glimcher, Shevach, and Paul in the mid 1970s, and I think it should be re-examined. I am virtually certain that the self-peptides that stimulate self-T cells in this assay will be the same self-peptides of which I speak; that is, self-peptides presented by self-MHC class II molecules to CD4 T cells that positively select in the thymus and contribute to “homeostatic” cycling in the periphery. However, what I would now like to address is the role of similar peptides in T cell activation. To examine this, we decided to cross our

5 Feb 2002

12:48

Annu. Rev. Immunol. 2002.20:1-28. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

22

AR

AR152-01.tex

AR152-01.SGM

LaTeX2e(2001/05/10)

P1: GSR

JANEWAY

TCR specific for the Eα52-68:I-Ab complex, called 1H3.1, with the mice kindly provided to us by Pippa Marrack and John Kappler. The initial crosses, not surprisingly, showed marked deletion of 1H3.1 T cells, as there were plentiful self-peptide:self-MHC class II complexes available on the cell surface. It was only when we got to mice that lacked genes for invariant chain and the Aβ b chain of I-Ab, and had the covalently linked Eα52-68:Aβ b and the 1H3.1 transgene, that we got a big surprise. Instead of intrathymic deletion, we observed intrathymic positive selection, At first, I was mystified by this result, but on a trip to Italy in the summer of 1999, I told Antonio Lanzavecchia about it, and suddenly all the pieces fell into place. The reason we saw positive selection could be that all of the TCRs in the “immunological synapse” between the two cells were binding to the same self-peptide (or foreign peptide) and undergoing the same conformational change, so that they could not form dimers. I was very excited by this result, which, after two years of sitting on it, we finally decided to submit to the Proceedings of the National Academies of Science. Christophe Viret was afraid that any number of artifacts could have contributed to the absence of responses and he tested for many of these, but I finally got him to give the paper to Pippa Marrack, and she has communicated it to the Proceedings (C. Viret, C. Janeway, 24). This paper, together with my inaugural article in the Proceedings, which makes several of the same points, should make Christophe Viret a high-profile immunologist when he returns to France. His manuscript is much more conservative than my own, but I feel I am old and experienced enough to get egg on my face if I am wrong.

A TRIP TO PHILADELPHIA One time, about five years ago, I went to Philadelphia to give a seminar. One member of the audience was an old friend of mine from our days at the NIH; his name is Mike Cancro. I actually thought that he had given up research for administration, but I was wrong; he had simply taken a deanship but was still pretty interested in and actively looking at B cell development. He really liked my seminar, and more than that, it had given him an idea that we were ideally placed to investigate. He wondered what would happen if we performed the same analysis on B cells as we had previously used to probe T cell development. I asked what we should look for, and he told me to compare κ light chain sequences in peripheral B lymphocytes that were either immature or mature in mice transgenic for a single heavy chain transgene. He pointed us in the right direction, telling me to get mice from Mark Shlomchik, but neither Mark nor I had a free set of hands. Therefore, I asked a very bright M.D., Ph.D. student named Matt Levine to take a look at one mouse to see if it was feasible. So Matt tried the first mouse he could get his hands on, one that bore a Vh186.2 heavy chain transgene and was homozygous for a complete knock-out of its JH region, so that no other heavy

5 Feb 2002

12:48

AR

AR152-01.tex

AR152-01.SGM

LaTeX2e(2001/05/10)

P1: GSR

Annu. Rev. Immunol. 2002.20:1-28. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

A TRIP THROUGH MY LIFE

23

chain could be formed. He sorted the cells based on the so-called E (immature) and F (mature) populations, extracted genomic DNA, and assayed light chain usage in the E and F populations. This required a great deal of sequencing, but it revealed one κ light chain that was five-fold higher in the mature population F as compared to the immature population E. He was encouraged by this result, so he did the same experiment using a different Ig-heavy chain transgene. This mouse yielded two different κ light chains that were expressed at about the same five-fold increase in population F as in population E, and they were different κ light chains than the one used by the first mouse. Matt eventually assayed seven mice, three for each heavy chain transgene, and one normal mouse. Each mouse showed the same difference in light chain usage, and so we wanted to publish our results in Immunity. However, we were blocked by a reviewer who wanted us to sequence both heavy and light chains on single cells. Finally, in frustration, we turned once again to Pippa Marrack. It turned out that her husband, John Kappler, had a slot for a communicated manuscript in PNAS, and so a paper appeared in the Proceedings of the National Academies of Science in 2000 entitled “Positive Selection in B Lymphocytes” (25). We included Mike Cancro as a couthor, as he basically told us what we needed to know, which was where to look and how. Having discovered positive selection in B cells, we then turned to the issue of what ligand(s) were doing the positive selecting? We wanted to test three hypotheses: The first was that they were influenced by the normal flora. To test this, we needed to get our mice made germ free. Thanks to an old friend named Ed Balish, whom I had met when I was looking at a job in Wisconsin, this turned out to be possible. We sent him several breeding pairs; he bred them and then delivered the pups by caesarian section. We assayed the κ light chain DNA, now by the more rapid technique of colony hybridization, and it turned out that germ-free mice showed the same positive selection that we had earlier seen in conventional mice. We next looked at whether prolonged expression of TdT, the enzyme that is used to synthesize nontemplated or N-nucleotides in the junctions between the V-D and D-J junctions, but only rarely to add nucleotides at Vκ -Jκ junctions, would affect positive selection. This led to the same results as the germ-free mice, but it took much longer to sort out because the Vκ - Jκ junctions were messed up by the enzyme. We next examined mice with the same heavy chain VH region as the mice used in our original experiment, but we engineered them in the cytoplasmic domain by removing the domain that permits secretion of the heavy chain. This gave us a big surprise. Mice that have only surface Ig but lack secreted Ig did not show positive selection, whereas our earlier results in the presence of secreted IgM showed clear signs of receptor-specific positive selection. Therefore, the ability to secrete Ig is essential for positive selection of the same Ig (unpublished data). This finding is consistent with Niels Jerne’s idiotypic network hypothesis (26), but it falls somewhat short of proving it. To prove that the idiotypic network is responsible for Matt Levine’s results, we need to try three experiments, all of

5 Feb 2002

12:48

Annu. Rev. Immunol. 2002.20:1-28. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

24

AR

AR152-01.tex

AR152-01.SGM

LaTeX2e(2001/05/10)

P1: GSR

JANEWAY

which are now underway. The first, and easiest, of these is to provide pooled immunoglobulin to mice that cannot secrete their heavy chain transgene. If this works, we then need to confirm it with serum from mice with the same heavy chain that can be secreted. The best, and most difficult, experiment is to breed an H-2 marker into the secretion competent or the secretion incompetent (but not both), make a mixed bone marrow chimera, and then harvest cells from the spleen when the mouse is old enough to show the effects of secreted Ig on the nonsecreting background. These experiments are all underway in the hands of a very enthusiastic group of young scientists, who are performing all these analyses at the same time. One problem is that all of our heavy chain transgenes have to be bred to JH −/− mice, but we are fortunate to be collaborating with Mark Shlomchik, our colleague at Yale, who has several B10 background mice that lack JH regions. These mice are ideal for moving a heavy chain transgene around from one strain to another, as the JH−/− mice are absolutely essential for this experiment to prevent endogenous heavy chain rearrangement. I am particularly happy to be doing this test of Niels Jerne’s elegant hypothesis. Many years ago, he said that I and my predecessor in this series, Herman Eisen, had both done experiments that confirmed the existence of the idiotypic network. We put our heads together, as neither one of us agreed with him. We then designed an experiment to test Niels Jerne’s hypothesis, but again, it involved tolerizing mice to idiotypes used as immunogens. As in all such experiments, we observed the synthesis of anti-idiotype antibody to purified idiotype, but that got us no closer to the mechanism by which this occurred (27). Years later, when Alexander (Sasha) Rudensky came to the lab, he wanted to make T cells that were specific for what he called an “idiopeptide,” but that would only tell us that such T cells existed and not what they do in ongoing T cell responses. I think that Matt Levine has discovered the true role of idiotypic networks, or at least a part of that role, in choosing the correct H:L pairs in the primary B cell receptor repertoire. That makes sense to me; to discover what mechanism is at work in such situations is a new project requiring new tools, but at least we have a hint that we can probe the network scientifically, and that is all I ask.

THE DEMONSTRATION BY JUAN LAFAILLE AND JEFF BLUESTONE THAT SUPPRESSOR CELLS EXIST The existence of suppressor T cells was initially postulated by the late Richard Gershon on the basis of two fundamental ideas: The first was that anything in biology that can go up also has to come down; he called this the second law of thymodynamics. The second was that unopposed stimulation would lead to autoimmune disease. These ideas have been validated in recent years, especially in experiments by Juan Lafaille and independently by Jeff Bluestone. Although these investigators prefer to call these cells regulatory T cells (Treg), this is really

5 Feb 2002

12:48

AR

AR152-01.tex

AR152-01.SGM

LaTeX2e(2001/05/10)

P1: GSR

Annu. Rev. Immunol. 2002.20:1-28. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

A TRIP THROUGH MY LIFE

25

a neologism for suppressor T cells, so I will stick with the original nomenclature throughout this analysis. Gershon was vindicated in a series of papers published over the last seven or eight years by Juan Lafaille, who was at that time working in Susumu Tonegawa’s lab at MIT. Using genes isolated from a clone of T cells that Jody Baron had prepared in my lab, called clone 19, which could produce a disease called experimental allergic encephalomyelitis (EAE), Juan prepared mice that were transgenic for the TCR of this clone. These mice could be induced to show EAE by stimulating the cells with the N-acetylated peptide 1-11 of myelin basic protein. To be sure that he had no endogenous genes for other TCRs, he crossed them to RAG−/−, I-Au mice, so this also required a great deal of breeding. Initially, they were breeding TCR transgenics with TCR transgenics, which explains a puzzling result in the first paper they published; the homozygous TCR transgenic mice got significant levels of spontaneous disease, much higher than in our own colony, where we bred only hemizygous mice. I really thought that they had infection in their colony, whereas our colony was specific pathogen free. It turned out later that they also did not observe disease in transgene hemizygous mice (28). Once they had mice that were TCR transgene positive, I-Au positive, and RAG-1−/−, they got a surprising result. The brain was flooded with MBP-specific T cells, and the mice succumbed to overwhelming EAE, even without immunization and in the absence of true signs of T cell activation. They published this result, we repeated it, and again saw overwhelming EAE. This result has been repeated many times, and always people see the same result. The real question was: What was missing in these mice that was present in the mice that could rearrange their T and B cell receptor genes? I heard Juan talk about this on many occasions, and he, and Susumu independently, later reported that what was missing was another kind of CD4 T cell that could suppress or protect the brain from attack (29, 30). One of the clear-cut messages of Juan Lafaille’s experiment was that the clone 19 TCR could not differentiate into a cell that suppresses the impetus of a clone that is autoreactive. Therefore, thanks to Juan’s effort to make an MBP TCR transgenic mouse that was bred onto a RAG-1 −/− mouse, we can infer the existence of the long-anticipated suppressor T cells. So you were right all along, Richard, and I personally apologize for years of my own skepticism. Having said that, it is not surprising that some T cells can escape from the thymus with autoreactive TCRs, but these are held in check by the existence of suppressor T cells. The question facing us now is whether these cells can be actively induced to make autoimmune disease even more unlikely. We believe that the answer is yes, as do several other laboratories who are working on the same question. So it appears that my thinking has come full circle, and that suppressor T cells do exist. We envision them as recognizing some self-peptide:self-MHC ligand as an agonist peptide, but secreting immunoregulatory cytokines such as TGFβ and IL-10. On the other hand, T cells will be activated to become autoaggressive in the case where a pathogen is involved, by putting CD80 and CD86 on the

5 Feb 2002

12:48

26

AR

AR152-01.tex

AR152-01.SGM

LaTeX2e(2001/05/10)

P1: GSR

JANEWAY

Annu. Rev. Immunol. 2002.20:1-28. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

antigen-presenting cell’s surface or by activating a dendritic cell that produces a cytokine profile such that Th1 cells are produced. So, in conclusion, all our T lymphocytes and all of our B lymphocytes are referential to self-ligands, yet they are not activated by such ligands unless they are presented on an activated antigen presenting cell. Even then, a T suppressor cell may hold them in check with the same or similar peptide and organ specificity. In this way, I envision that suppressor T cells generally are dominant over autoaggressive T cells and thereby confine the number of autoimmune diseases to certain organs that are protected by other means, such as the blood-brain barrier in the case of EAE or in multiple sclerosis for which EAE serves as a model.

SUMMARY In conclusion, I want young scientists to continue up these same paths, and others, in studying how the immune system operates. I can imagine wonderful discoveries in the future, none of which is as important as discovering how HIV-1 produces AIDS and the devising of an effective vaccine against this deadly pathogen. In this essay, I have attempted to chart out the meanders of my own career, as an exemplar of how to conduct research, and also, where the results were wrong, how not to conduct research. I have tried to make four main points. First, that the immune system exists to protect the body from infection, and it does so in the first instance by using an ancient system of host defense. Second, that the adaptive immune system is referential to self-ligands, that is, T and B lymphocytes are selected on self-ligands, sustained in the periphery by self-ligands, and use recognition of self-ligands in mounting responses to foreign antigens. Third, never assume you know all things, as I did about autoimmune disease; always push that last mile, as Juan Lafaille did. And fourth, be inspired by the knowledge that exists at the time you enter research, but be irreverent toward this knowledge like my friend Ian McConnell, for this is the road to true understanding. ACKNOWLEDGMENTS First and foremost, I would like to acknowledge my great debt to my wife, Kim Bottomly, who, while working on Th1 and Th2 generation and function, took the time to care for my daughter Katie and our daughters Hannah and Megan. She also helped me in preparing my course notes, took care of me when I was sick, and generally gave everything that one scientist can give another scientist. Second, I would like to thank Jennifer Boucher-Reid for typing the same manuscript over and over again, and not complaining about it. Third, I would like to thank my whole laboratory, past, present, and I hope future for making my 25+ years at Yale so exciting and informative. Lastly, I would like to thank all the students I have taught at Yale Medical School.

5 Feb 2002

12:48

AR

AR152-01.tex

AR152-01.SGM

LaTeX2e(2001/05/10)

P1: GSR

A TRIP THROUGH MY LIFE

27

Visit the Annual Reviews home page at www.annualreviews.org

Annu. Rev. Immunol. 2002.20:1-28. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

LITERATURE CITED 1. Padovan E, Mauri-Hellweg D, Pichler WJ, Weltzien HU. 1996. T cell recognition of penicillin G: structural features determining antigenic specificity. Eur. J. Immunol. 26:42–48 2. Agrawal A, Eastman QM, Schatz DG. 1998. Transposition mediated by RAG1 and RAG2 and its implications for the evolution of the immune system. Nature 394:744–51 3. Liu Y, Janeway CA Jr. 1992. Cells that present both specific ligand and costimulatory activity are the most efficient inducers of clonal expansion of normal CD4 T cells. Proc. Natl. Acad. Sci. USA 89:3845–49 4. Galvin F, Freeman GJ, Razi-Wolf Z, Hall W Jr, Benacerraf B, Nadler L, Reiser H. 1992. Murine B7 antigen provides a sufficient costimulatory signal for antigenspecific and MHC-restricted T cell activation. J. Immunol. 149:3802–8 5. Liu Y, Janeway CA Jr. 1991. Microbial induction of co-stimulatory activity for CD4 T-cell growth. Int. Immunol. 3:323–32 6. Janeway CA Jr. 1989. Approaching the asymptote? Evolution and revolution in immunology. Cold Spring Harb. Symp. Quant. Biol. 54:1–13 7. Medzhitov R, Preston-Hurlburt P, Janeway CA Jr. 1997. A human homologue of the Drosophila Toll protein signals activation of adaptive immunity. Nature 388:394– 97 8. Lemaitre B, Nicolas E, Michaut L, Reichhart JM, Hoffmann JA. 1996. The dorsoventral regulatory gene cassette spatzle/ Toll/cactus controls the potent antifungal response in Drosophila adults. Cell 86: 973–83 9. Takeuchi O, Hoshino K, Kawai T, Sanjo H, Takada H, Ogawa T, Takeda K, Akira S. 1999. Differential roles of TLR2 and TLR4 in recognition of gram-negative and gram-

10.

11.

12. 13.

14.

15.

16. 17. 18.

19.

20.

positive bacterial cell wall components. Immunity 11:443–51 Hayashi F, Smith KD, Ozinsky A, Hawn TR, Yi EC, Goodlett DR, Eng JK, Akira S, Underhill DM, Aderem A. 2001. The innate immune response to bacterial flagellin is mediated by Toll- like receptor 5. Nature 410:1099–103 Hemmi H, Takeuchi O, Kawai T, Kaisho T, Sato S, Sanjo H, Matsumoto M, Hoshino K, Wagner H, Takeda K, Akira S. 2000. A Toll-like receptor recognizes bacterial DNA. Nature 408:740–45 Medzhitov R, Janeway C Jr. 2000. Innate immunity. N. Engl. J. Med. 343:338–44 Ayehunie S, Garcia-Zepeda EA, Hoxie JA, Horuk R, Kupper TS, Luster AD, Ruprecht RM. 1997. Human immunodeficiency virus-1 entry into purified blood dendritic cells through CC and CXC chemokine coreceptors. Blood 90:1379–86 Burnet FM. 1959. The Clonal Selection Theory. London: Cambridge Univ. Press. 209 pp. Gershon RK, Kondo K. 1971. Infectious immunological tolerance. Immunology 21: 903–14 von Boehmer H. 1994. Positive selection of lymphocytes. Cell 76:219–28 Nossal GJ. 1994. Negative selection of lymphocytes. Cell 76:229–39 Viret C, Wong FS, Janeway CA Jr. 1999. Designing and maintaining the mature TCR repertoire: the continuum of self-peptide:self-MHC complex recognition. Immunity 10:559–68 Brocker T. 1997. Survival of mature CD4 T lymphocytes is dependent on major histocompatibility complex class II-expressing dendritic cells. J. Exp. Med. 186:1223–32 Tanchot C, Lemonnier FA, Perarnau B, Freitas AA, Rocha B. 1997. Differential requirements for survival and proliferation

5 Feb 2002

12:48

28

21.

Annu. Rev. Immunol. 2002.20:1-28. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

22.

23.

24.

25.

26.

AR

AR152-01.tex

AR152-01.SGM

LaTeX2e(2001/05/10)

P1: GSR

JANEWAY of CD8 naive or memory T cells. Science 276:2057–62 Ignatowicz L, Kappler J, Marrack P. 1996. The repertoire of T cells shaped by a single MHC/peptide ligand. Cell 84:521–29 Hunt DF, Michel H, Dickinson TA, Shabanowitz J, Cox AL, Sakaguchi K, Appella E, Grey HM, Sette A. 1992. Peptides presented to the immune system by the murine class II major histocompatibility complex molecule I-Ad. Science 256:1817–20 Alam SM, Davies GM, Lin CM, Zal T, Nasholds W, Jameson SC, Hogquist KA, Gascoigne NR, Travers PJ. 1999. Qualitative and quantitative differences in T cell receptor binding of agonist and antagonist ligands. Immunity 10:227–37 Viret C, He X, Janeway CA Jr. 2001. Paradoxical intrathymic selection in mice with only a covalently presented agonist peptide. Proc. Natl. Acad. Sci. USA. 98:9243–48 Levine MH, Haberman AM, Sant’Angelo DB, Hannum LG, Cancro MP, Janeway CA Jr, Shlomchik MJ. 2000. A B-cell receptor-specific selection step governs immature to mature B cell differentiation. Proc. Natl. Acad. Sci. USA 97:2743–48 Jerne NK. 1974. Towards a network the-

27.

28.

29.

30.

ory of the immune system. Ann. Immunol. (Paris) 125C:373–89 Sakato N, Janeway CA Jr, Eisen HN. 1977. Immune responses of BALB/c mice to the idiotype of T15 and of other myeloma proteins of BALB/c origin: implications for an immune network and antibody multispecificity. Cold Spring Harb. Symp. Quant. Biol. 41:719–24 Lafaille JJ, Nagashima K, Katsuki M, Tonegawa S. 1994. High incidence of spontaneous autoimmune encephalomyelitis in immunodeficient anti-myelin basic protein T cell receptor transgenic mice. Cell 78:399–408 Olivares-Villagomez D, Wang Y, Lafaille JJ. 1998. Regulatory CD4(+) T cells expressing endogenous T cell receptor chains protect myelin basic protein-specific transgenic mice from spontaneous autoimmune encephalomyelitis. J. Exp. Med. 188:1883– 94 Van de Keere F, Tonegawa S. 1998. CD4(+) T cells prevent spontaneous experimental autoimmune encephalomyelitis in anti-myelin basic protein T cell receptor transgenic mice. J. Exp. Med. 188:1875– 82

8 Feb 2002

9:31

AR

AR152-02.tex

AR152-02.SGM

LaTeX2e(2001/05/10) P1: GSR 10.1146/annurev.immunol.20.091101.091806

Annu. Rev. Immunol. 2002. 20:29–53 DOI: 10.1146/annurev.immunol.20.091101.091806 c 2002 by Annual Reviews. All rights reserved Copyright °

THE B7 FAMILY OF LIGANDS AND ITS RECEPTORS:

Annu. Rev. Immunol. 2002.20:29-53. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

New Pathways for Costimulation and Inhibition of Immune Responses Beatriz M. Carreno and Mary Collins Genetics Institute/Wyeth Research, 87 Cambridge Park Drive, Cambridge, Massachusetts 02140; e-mail: [email protected]; [email protected]

Key Words ICOS, PD-1, PD-L, T cell activation, tolerance ■ Abstract T cell activation is dependent upon signals delivered through the antigen-specific T cell receptor and accessory receptors on the T cell. A primary costimulatory signal is delivered through the CD28 receptor after engagement of its ligands, B7-1 (CD80) or B7-2 (CD86). Engagement of CTLA-4 (CD152) by the same B7-1 or B7-2 ligands results in attenuation of T cells responses. Recently, molecular homologs of CD28 and CTLA-4 receptors and their B7-like ligands have been identified. ICOS is a CD28-like costimulatory receptor with a unique B7-like ligand. PD-1 is an inhibitory receptor, with two B7-like ligands. Additional members of B7 and CD28 gene families have been proposed. Integration of signals through this family of costimulatory and inhibitory receptors and their ligands is critical for activation of immune responses and tolerance. Understanding these pathways will allow development of new strategies for therapeutic intervention in immune-mediated diseases.

THE B7/CD28/CTLA-4 PATHWAY: THE PARADIGM FOR THE FAMILY The CD28/CTLA-4/B7-1/B7-2 family provides a paradigm with which to define new related immune pathways. From this pathway, we find that multiple B7 ligands bind to both activating (CD28) and inhibitory (CTLA-4) receptors (1–3). These receptors do not function independently, but they modify responses delivered by engagement of the antigen-specific TCR on T cells. To date, members of the receptor family are type I transmembrane proteins with a single IgV extracellular domain, and the ligands are type I transmembrane proteins with both IgV and IgC extracellular domains. Interactions between the receptor-ligand pairs are mediated predominantly by residues in the IgV domains. Expression of both receptors and ligands is tightly regulated, allowing discrimination between signals that result in activation or inhibition of an immune response. The CD28 receptor is constitutively expressed on T cells. Engagement of CD28 on naive T cells by either B7-1 or B7-2 ligands on antigen-presenting cells 0732-0582/02/0407-0029$14.00

29

4 Feb 2002

11:14

Annu. Rev. Immunol. 2002.20:29-53. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

30

AR

AR152-02.tex

CARRENO

¥

AR152-02.SGM

LaTeX2e(2001/05/10)

P1: GSR

COLLINS

provides a potent costimulatory signal to T cells activated through their T cell receptor (1–3). This results in induction of IL-2 transcription, expression of CD25, and entry into the cell cycle. CD28 engagement also confers critical survival signals to T cells through the Bcl-XL pathway (4). CD28 costimulation is necessary for the initiation of most T cell responses, and blockade of CD28 signaling results in ineffective T cell activation. This has therapeutic implications, in that blockade of CD28 costimulation can be profoundly immunosuppressive, preventing induction of pathogenic T cell responses in autoimmune disease models and allowing for prolonged acceptance of allografts in models of organ transplantation (1, 2). CTLA-4 (CD152) shares about 30% identity with CD28 at the amino acid level (Table 1). CTLA-4 expression is not detected on na¨ıve T cells but is transcriptionally induced after T cell activation (3). Cell surface expression of CTLA-4 is very tightly regulated, with most of the CTLA-4 protein residing within cytoplasmic vesicles (5). The critical role of CTLA-4 as a negative regulator of T cell activation is dramatically illustrated in CTLA-4-deficient mice, which die within 3 to 4 weeks of birth from massive lymphocytic infiltration and tissue destruction in critical organs (6–8). Both CD28 and CTLA-4 share binding to B7-1 (CD80) and B7-2 (CD86) ligands. B7-1 and B7-2 are capable of forming homodimers, allowing for interactions with homodimers of either CTLA-4 or CD28. The interaction of CD28 with its ligands is weaker than the interaction with CTLA-4. Human B7-1 binds to human CTLA-4 and CD28 with Kd values of 0.42 and 4 µM, respectively for the monomeric interactions (9). The B7-2:CTLA-4 interaction is of an affinity similar to B7-1:CD28, and the CD28:B7-2 interaction is of even lower affinity (10). Mice

TABLE 1 B7 family receptors: Amino acid identities (%) were calculated using the Wisconsin Package (GCG) Version 10 Gap program, and the Blosum 62 scoring matrix. For sequence comparisons indicated with an asterisk, the Structgappep scoring matrix was used, as the sequences were too distantly related to align with Blosum 62. Comparisons between all mouse (m) and human (h) proteins are shown Protein identities (%) among B7 family receptors Protein

mCD28

hCD28

69

mCD28 hCTLA-4 mCTLA-4 hICOS mICOS hPD-1

hCTLA-4

mCTLA-4

hICOS

mICOS

hPD-1

mPD-1

33

34

29

29

15*

18*

30

32

24

25

12*

20*

74

18

18

20

18*

21

18

22

18*

69

13*

16*

12*

14* 60

4 Feb 2002

11:14

AR

AR152-02.tex

AR152-02.SGM

LaTeX2e(2001/05/10)

P1: GSR

Annu. Rev. Immunol. 2002.20:29-53. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

NEW B7-LIKE LIGANDS AND RECEPTORS

31

deficient in B7-1 and B7-2 have significant abnormalities in both humoral and cellular immune responses, again illustrating the fundamental role of this pathway (11). The recent crystal structures of CTLA-4-B7 complexes are characterized by homodimers of CTLA-4 that contain B7-binding sites located distally to the CTLA-4 dimer interface (12, 13). The crystal structures suggest that the CTLA-4 homodimer can bind to noncovalent homodimers of B7-1 or B7-2 to form a lattice of CTLA-4-B7 interactions (12, 13). Formation of such a lattice could function to form a stable signaling complex at the T cell–APC interface. CD28 also forms homodimers, with a conserved cysteine located just proximal to the transmembrane domain linking the monomers in CD28 and CTLA-4. By homology, CD28 may also form lattice structures with B7-1 and B7-2, which could serve to potentiate the costimulatory signals delivered through CD28. With the determination of the human genome sequence, and with significant accumulation of mouse genomic sequences, algorithms for gene homology can be used to identify genes encoding proteins with structural homology to B7 and CD28/CTLA-4. The Ig superfamily represents a large number of proteins, including BCR and TCR, and thus it does not provide sufficient criteria to identify a protein as a new costimulatory or inhibitory receptor or ligand. Instead, proteins with the highest homology to known gene family members are identified and tested for functional relationships. In addition, gene mapping can suggest evolutionary relationships. Recently, several proteins have been identified as new members of the B7 and CD28/CTLA-4 families (Figure 1). The B7 family ligands have been selected based on identities of about 20% to 30% in the extracellular domains, and they are characterized by an amino terminal signal peptide, one Ig-V and one Ig-C domain, a transmembrane domain, and a cytoplasmic tail. Inclusion in the B7 family of ligands is based upon degree of homology, as well as on evidence of costimulatory or inhibitory function in immune assays. New transmembrane proteins related to the CD28 receptor family have been identified based upon homology, IgV extracellular domain structure, functional activity, and binding to a B7-like ligand.

THE ICOS PATHWAY: IDENTIFICATION OF ICOS AND ITS LIGAND ICOS (AILIM) is a costimulatory receptor homologous to CD28 and CTLA-4 (14, 15) (Table 1). Human ICOS is a 55–60 kDa, disulfide-linked, glycosylated homodimer when isolated from activated human T cells. The protein has two putative N-glycosylation sites, and the unglycosylated monomer has a molecular weight of about 20 kDa (16). Mouse ICOS is a 47–57 kDa, disulfide-linked, N-glycosylated homodimer (17). Rat ICOS is highly homologous (85% identity) to the mouse protein (18). In rat, a differentially spliced form of ICOS containing a longer cytoplasmic tail has been identified, which suggests that additional forms may exist in human and mouse (18).

4 Feb 2002

11:14

Annu. Rev. Immunol. 2002.20:29-53. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

32

AR

AR152-02.tex

CARRENO

¥

AR152-02.SGM

LaTeX2e(2001/05/10)

P1: GSR

COLLINS

Figure 1 Summary of B7 family ligands and their receptors: The names of receptors and ligands are indicated, as well as a brief summary of predominant expression patterns for each. The conserved structure of a single IgV extracellular domain for receptors and IgV and IgC extracellular domains for ligands is depicted at the top. Function arrows summarize whether the pathway is thought predominantly to costimulate or inhibit the response of the receptor-bearing cell.

ICOS lacks the canonical MYPPPY motif, which is present in the extracellular domains of CD28 and CTLA-4, but contains a related FDPPPF sequence in the analogous position in the protein. The crystal structures of the CTLA-4/B7-1 and CTLA-4/B7-2 complexes implicate the MYPPPY site as the major contact site in CTLA-4 with B7-1 and B7-2 (12, 13). Although the structure of CD28 has not yet been solved, amino acid homologies, mutation data, and modeling support the concept that this motif will also be a major B7-binding site for CD28 (19). The related FDPPPF site in ICOS is not sufficiently conserved to allow for detectable binding of ICOS to B7-1 or B7-2 (16, 20, 21). However, structural homology raises the possibility that this motif might be important in the binding of ICOS to its ligand, which is a member of the B7 family (see below). The ICOS gene is closely linked to the genes for CD28 and CTLA-4 on human chromosome 2q33 and mouse chromosome 1 (17, 22). In humans, the three genes form a tightly linked cluster, with a gene order of CD28-CTLA-4-ICOS within a 300-kb region (22a), suggesting that these genes originated by gene duplication. This tight clustering suggests that expression of these genes may be coordinately

4 Feb 2002

11:14

AR

AR152-02.tex

AR152-02.SGM

LaTeX2e(2001/05/10)

P1: GSR

Annu. Rev. Immunol. 2002.20:29-53. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

NEW B7-LIKE LIGANDS AND RECEPTORS

33

regulated, with alterations in chromatin structure affecting the entire locus, as observed for tightly linked cytokine genes (23). Moreover, disease susceptibility loci mapped to this region must include ICOS in addition to CD28 and CTLA-4 as candidate genes. ICOS is an inducible costimulatory receptor expressed on activated, but not resting, T cells (14); ICOS is identical to the H4 T cell activation antigen (24). Expression of ICOS on T cells is dependent upon both TCR and CD28 signals, in that activation of T cells in the absence of CD28 engagement results in diminished levels of ICOS on T cells (25). This suggests that signals delivered by ICOS will typically occur distally to those delivered through CD28. However, ICOS expression is not absolutely dependent upon CD28 signals because activated human CD8+ T cells that do not express CD28 can express ICOS (14), and some T cell responses in CD28-deficient mice can be modulated with ICOS.Fc (26). Interestingly, although blockade of the CD28-B7 interaction opposes induction of ICOS expression, blockade of the CD40-CD40L pathway has no effect on ICOS induction on activated human T cells (16). The costimulatory functions of both CD28 and ICOS raise the possibility that they may share overlapping signaling pathways. The cytoplasmic tail of CD28 contains a YMNM motif, which is a PI-3 kinase binding site, and which can bind Grb2 and a Grb2-related protein GADS/GRID (27). Mutations in this site result in a failure to effectively recruit PI3K and a failure to induce Bcl-XL expression in response to CD28 ligation (27, 28). In addition, a consensus SH3-kinase binding site, PYAP, distal to the YMNM site is critical for costimulation of proliferation and IL-2 production (28, 29). The cytoplasmic tail of ICOS contains a YMFM motif, which binds the p85 subunit of PI3K, although binding of Grb-2 to ICOS was not detected (22). The cytoplasmic tail of ICOS lacks the PXXP site implicated in IL-2 production by CD28 engagement, which may account, in part, for the distinct functions of CD28 and ICOS. ICOS is expressed on T cells in lymphoid organs, such as spleen, lymph node, and Peyer’s patches in human and mouse (14, 16, 17, 21). ICOS+ T cells are found in germinal centers and surrounding T cell zones, and ICOS expression in these areas is enhanced after immune priming (21). CXCR5+ T cells, a subset of CD4+ T cells that are found in B cell follicles and germinal centers, are highly enriched in ICOS expression, as compared with T cells from peripheral blood (30, 31). This activated T cell subset is likely to be involved in enhancement of antibody responses. ICOS is expressed in the medulla and the cortico-medullary junction of the thymus (17). However, mice deficient for ICOS have a normal thymus and normal numbers of peripheral CD4+ and CD8+ T cells, suggesting that ICOS does not play a critical role in T cell development (32–34). In human, ICOS expression was detected in fetal and newborn thymuses, with expression primarily in the medulla (16). These data suggest that ICOS could contribute to thymic development, but analysis of ICOS-deficient mice indicates that ICOS is not obligatory. Similarly, mice deficient in CD28 have normal thymuses and normal numbers of peripheral

4 Feb 2002

11:14

34

AR

AR152-02.tex

CARRENO

¥

AR152-02.SGM

LaTeX2e(2001/05/10)

P1: GSR

COLLINS

TABLE 2 B7 family ligands: Amino acid identities (%) were calculated using the Wisconsin Package (GCG) Version 10 Gap program, and the Blosum 62 scoring matrix. Comparisons between all mouse (m) and human (h) proteins are shown. At the time of this review, the sequence for mouse B7-H3 was not available Protein identities (%) among B7 family ligands Protein

mB7.1

hB7.2

mB7.2

hICOS-L

mICOS-L

hB7H3

hPD-L1

mPD-L1

hPD-L2

mPD-L2

hB7.1

45

26

24

26

26

27

25

25

23

24

30

27

24

27

26

26

25

24

25

51

23

27

26

21

26

22

20

25

25

28

24

24

20

21

48

33

25

27

27

28

31

25

25

22

29

31

28

27

26

70

41

39

43

38

mB7.1 hB7.2

Annu. Rev. Immunol. 2002.20:29-53. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

mB7.2 hICOS-L mICOS-L hB7H3 hPD-L1 mPD-L1 hPD-L2

70

T and B cells, indicating that these costimulatory pathways are not obligatory for normal T cell development (35). The ligand for ICOS (B7h, GL50, B7RP-1, LICOS, B7-H2, KIAA0653), which is called ICOS-L for clarity and nonpartisanship in this review, was identified as a B7-like molecule capable of binding to ICOS and delivering a costimulatory signal to T cells (17, 20, 21, 36–38). Comparisons indicate that mouse and human ICOS-L share about 48% amino acid identity, and about 25% identity with other members of the B7 family (Table 2). Human ICOS-L maps to chromosome 21q22.3 and has been annotated on the chromosome 21 DNA sequence at position 31156109 (39, 40). The location of the human gene suggests that mouse ICOS-L should map in the syntenic region of mouse chromosome 10. Thus, in contrast to the CD28/CTLA-4/ICOS gene cluster, ICOS-L is in a distinct location from the B7-1 and B7-2 genes, which are on human chromosome 3q13.3-21 and mouse chromosome 16, respectively (41). Interestingly, the transcripts identified by the different groups as ICOS-L are not identical but result from differential splicing patterns (42). Two predicted forms of the human protein have been identified from immune sources, encoding proteins with differences in their cytoplasmic tails. It will be interesting to determine whether there are functional implications for this differential splicing. Differential splice variants have also been observed for the B7-1 and B7-2 genes (43–45). Measurement of the binding affinity of ICOS to ICOS-L indicates that the affinity of this interaction is very comparable to that of CD28 and B7-1 (21, 38). Davis and colleagues (38) measured a Kd of 4 µM for the interaction of monomers of ICOS-L with immobilized ICOS.Fc at 37◦ C, similar to their estimates of the affinities for B7-1 and CD28 (9, 38). Although they detected weak binding of tetramers of ICOS-L to CD28 and CTLA-4 at 25◦ C, there was no binding of monomers at 37◦ C; other groups have failed to detect this association using dimeric

4 Feb 2002

11:14

AR

AR152-02.tex

AR152-02.SGM

LaTeX2e(2001/05/10)

P1: GSR

Annu. Rev. Immunol. 2002.20:29-53. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

NEW B7-LIKE LIGANDS AND RECEPTORS

35

reagents. This suggests that ICOS-L is not a physiological ligand for CD28 or CTLA-4. ICOS-L is expressed constituitively on unstimulated B cells, splenic and peritoneal macrophages, and peripheral blood–derived dendritic cells by binding of ICOS.Fc to murine cells (20, 21, 36). It has also been detected on a small subset of CD3+ T cells, like other B7-family members (20). In humans, ICOS.Fc bound to peripheral B cells from some donors, and to monocytes and monocytederived dendritic cells (46). Interestingly, INFγ , but not LPS or TNFα or antiCD40L, augmented the expression of ICOS-L on CD14+ monocytes. ICOS-L expression has been detected by analysis of mRNA in many nonlymphoid tissues, such as kidney, liver, heart, and brain (20, 36). TNFα further induces expression of ICOS-L on B cells and monocytes, and most interestingly, it induces the expression of ICOS-L on fibroblasts by mRNA analysis. LPS treatment of mice, which induces TNF-α production, induced ICOS-L mRNA expression in testes, kidney, and peritoneum (36). ICOS-L expression was downmodulated in spleen cells from mice treated with LPS. These data suggest that ICOS-L may be induced by inflammatory signals in peripheral sites, although this must be confirmed at the protein level. Examination of ICOS-L protein in rejecting heart transplants showed that ICOS-L protein was expressed by interstitial dendritic cells in normal myocardium, and additionally on large inflammatory macrophages in rejecting hearts (47).

MODULATION OF IMMUNE RESPONSES BY THE ICOS PATHWAY Engagement of ICOS on T cells that have been stimulated through the TCR results in augmented proliferative responses and cytokine production (14). In comparisons of costimulation mediated through CD28 and ICOS, production of IL-2 is most effectively induced by CD28 (21, 48, 49), although modest enhancement of IL-2 production by ICOS engagement has been reported (37, 50). Costimulation of human CD4+ T cells by ICOS does not produce sustained proliferative responses due to limiting IL-2 production (50). Costimulation through ICOS is particularly effective in enhancing IL-10 production; in direct comparisons, ICOS is more potent than CD28 in inducing the production of IL-10 (14, 16, 25, 37). Both pathways augment the production of other effector cytokines such as IFNγ , IL-4, IL-5, and TNFα (25, 49, 50). Thus, CD28 costimulation appears to have a nonredundant role in the initial costimulation of IL-2 and is critical for initiation of immune responses. In contrast, the subsequent expression of ICOS and engagement by ICOS-L is more important for augmentation of IL-10 and enhancement of effector functions. ICOS engagement can augment induction of both Th1 and Th2 cytokines, but under some circumstances it may more effectively costimulate Th2 responses. ICOS is expressed similarly on both Th1 and Th2 lines after primary stimulation but remains high only on Th2 lines after repeated activation steps (22, 25). This

4 Feb 2002

11:14

Annu. Rev. Immunol. 2002.20:29-53. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

36

AR

AR152-02.tex

CARRENO

¥

AR152-02.SGM

LaTeX2e(2001/05/10)

P1: GSR

COLLINS

suggests that blockade of the ICOS pathway could affect both primary Th1 and Th2 responses, but that highly polarized Th2 responses might be more affected by ICOS blockade. Kopf et al. (26) show that both the Th1 cytokine IFNγ and Th2 cytokines IL-4 and IL-5 are reduced by administration of ICOS.Fc at the time of infection with N. brasiliensis, indicating that the ICOS pathway can enhance production of both Th1 and Th2 cytokines in vivo. In addition, Coyle et al. (22) show that production of both Th1 and Th2 cytokines can be suppressed by addition of ICOS.Fc at the time of reactivation in vitro. However, for highly polarized lines, only the production of Th2 cytokines is reduced by ICOS blockade (22). In addition, adoptive transfer of polarized Th2 lines in a lung inflammatory model results in a dependence upon ICOS engagement in vivo for optimal production of Th2 cytokines and resulting eosinophilia. The highly polarized Th1 line, which induces a neutrophilic infiltrate, is not ICOS dependent (22). Thus, it appears that Th2 cytokine production will generally be more dependent upon ICOS costimulation, whereas the dependence upon ICOS for Th1 cytokines will be determined by the precise conditions under which that Th1 response is elicited. Recent studies in EAE suggest that ICOS costimulation may play a larger role in the effector phase of a Th1 response, in that disease is ameliorated by blockade of ICOS only during the effector phase (51). CD28 costimulation augments production of both Th1 and Th2 cytokines as well, but Th2 responses are also more dependent upon CD28 costimulation (1). Thus, it may be that Th2 responses simply require a higher threshold of costimulatory signals, and that both CD28 and ICOS contribute to the induction of Th2 responses. Interestingly, CTLA-4 engagement can oppose T cell activation with costimulatory signals delivered by either CD28 or ICOS (50), and thus the outcome of the immune response will also depend upon whether CTLA-4 is engaged concomitantly with ICOS by APC that express both B7 ligands and ICOS-L. ICOS engagement can also influence CD8+ T cell responses. Expression of ICOS-L in an immunogenic, MHC class I+ tumor resulted in enhanced tumor rejection in mice (52). In these studies, ICOS-L costimulation of CD8+ T cells was found to enhance IL-2 and IFN-γ production preferentially in recall responses compared with naive responses. No enhancement of CD8+ lytic effector function was observed (52), which is consistent with studies showing that inhibition of the ICOS pathway had no effect on CTL responses after LCMV or VSV infection in mice (26). Thus, generation of lytic effector functions in CD8+ T cells does not appear to be ICOS-dependent. The ICOS pathway appears to play a large role in antibody responses and germinal center formation. ICOS is expressed by germinal center T cells, and its ligand is expressed by splenic B cells (21). Transgenic mice expressing a secreted form of ICOS-L.Fc protein are characterized by lymphoid hyperplasia in the spleen, lymph nodes, and Peyer’s patches, and have high serum levels of IgG (21). Evaluation of ICOS-deficient mice by three independent groups supports a critical role for ICOS in humoral immunity (32–34). ICOS-deficient mice have a consistent decrease in serum IgG1 levels (32, 34), and immunization of mice with TNP-KLH in the absence of adjuvant or with alum or IFA reveals a deficit in IgG1 and IgG2a antibody

4 Feb 2002

11:14

AR

AR152-02.tex

AR152-02.SGM

LaTeX2e(2001/05/10)

P1: GSR

Annu. Rev. Immunol. 2002.20:29-53. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

NEW B7-LIKE LIGANDS AND RECEPTORS

37

production (32). This deficit could be overcome by the use of CFA as an adjuvant (32). However, in another study, immunization with KLH in CFA also resulted in decreased isotype switching (34), suggesting that the precise immunization conditions and antigen used may affect the outcome. Immunization of mice with NP-OVA in alum (34) or with aerosolized antigen in the lung (33) also revealed a deficit in IgE production in ICOS-deficient mice. Thus, under several conditions of immunization, deficits in isotype switching have been observed. This deficit is not rescued by secondary immunization and appears to be due to a lack of T cell help, as isotype switching to T cell–independent antigens is intact (34). Activation of CD40 was able to rescue the defect in isotype switching in ICOS-deficient mice (32). Engagement of ICOS can enhance anti-CD3-mediated induction of CD40L on T cells, indicating that both CD28 and ICOS can contribute to activation of the CD40L-CD40 pathway. ICOS-L is highly expressed in the B cell–rich areas of the spleen (21), consistent with the proposal that activation of the CD40 pathway may be the critical event mediated by the ICOS pathway in the development of a humoral response (32). However, expression of CD40L is clearly not absolutely dependent upon ICOS expression, as ConA-activated T cells from ICOS-deficient mice can express normal levels of CD40L (53). Consistent defects in germinal center formation are also observed in ICOSdeficient mice (32–34). Mice form fewer and smaller germinal centers in response to both primary and secondary immunization (32–34). Mice deficient in CD28 (35) or in both B7-1 and B7-2 (11) are also defective in isotype switching and germinal center formation. Similarly, mice deficient in CD40L have severe deficits in isotype switching and germinal center formation (54, 55). It will be interesting to determine precisely how these three pathways intersect in directing the humoral immune response.

THERAPEUTIC IMPLICATIONS: THE ICOS PATHWAY The apparent increased dependence of Th2 responses on the ICOS pathway and the clear role for ICOS in isotype switching raise the possibility that targeting this pathway may be useful in generation of therapeutics for diseases with antibodymediated and Th2-mediated pathologies. Interestingly, antagonism of ICOS appears to be more effective late in an immune response. Blockade of ICOS at the time of antigen priming for lung inflammation had little effect upon subsequent airway challenge in normal mice (49). In addition, ICOS-deficient mice are still susceptible to induction of inflammatory lung disease induced by airway challenge with OVA in primed mice (33). The absence of ICOS in this model results in lower production of IL-4 and IL-13, but no change in the lung histology (33). In contrast, antagonism of ICOS 21 days after priming significantly reduced lung inflammation after airway antigen challenge (49). Similarly, ICOS-blockade decreased lung inflammation and airway hyperreactivity after adoptive transfer of highly polarized Th2 cells to naive mice (22). Blockade of the B7/CD28 pathway

4 Feb 2002

11:14

Annu. Rev. Immunol. 2002.20:29-53. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

38

AR

AR152-02.tex

CARRENO

¥

AR152-02.SGM

LaTeX2e(2001/05/10)

P1: GSR

COLLINS

also prevents airway hyperreactivity after adoptive transfer of either Th1 or Th2 cells lines and, further, does not show the Th2-bias of the ICOS blockade. Antagonism of CD28 is more effective at the time of antigen priming and less effective at later times in these models of lung inflammation (49). This suggests that the major contributions of CD28 and ICOS in costimulation of Th2 responses occur at different times during generation of this response. Induction of EAE, a Th1-mediated autoimmune disease, is not dependent upon ICOS, in that antagonism of ICOS at the time of antigen priming does not prevent disease (51). Surprisingly, ICOS-deficient mice are more susceptible to EAE (33), and blockade of ICOS at the time of priming for EAE results in more severe disease (51). The enhanced disease could be due to the absence of a protective Th2 response, as a deficiency in IL-13 production was noted (33). Other explanations are possible at this time, including defects in IL-10 production or defects in regulatory cells, both of which would be expected to result in enhanced disease (56, 57). CD4+ CD25+ regulatory cells are dependent upon CD28 for maintenance in the periphery (58) but have not yet been examined in ICOS-deficient mice. ICOS costimulation in EAE does appear to be critical at the time that encephalitogenic T cells begin to migrate into the CNS (51). Antagonism of ICOS at this time reduced disease severity. Mice deficient in CD28 or in both B7-1 and B7-2 are resistant to EAE (59). Interestingly, adoptive transfer of primed encephalitogenic T cells into B7-double deficient mice also results in reduced disease, implicating CD28 engagement in both priming and effector stages of the disease (59). The ICOS pathway plays a role in graft rejection, in that blockade of the ICOS pathway results in prolongation of heart allograft survival in mouse models (47). Combination of anti-ICOS antibody and anti-CD40L antibody in this heart transplant model also reduced vasculopathy in the cardiac grafts compared with antiCD40L alone, suggesting that the ICOS pathway is contributing to chronic allograft rejection (47). Similarly, combinations of agents that block the B7/CD28 pathway and anti-CD40L promote long-term cardiac allograft survival in mice and prevent development of vascular lesions associated with chronic rejection (60). These data suggest that both ICOS and CD28 are contributing to inflammatory stimuli underlying the chronic rejection pathology. IL-4 production has been linked to development of transplant arteriosclerosis in mice deficient for CD40 (61). It is possible that both CD28 and ICOS contribute to chronic rejection pathology by costimulation of Th2 responses that could accompany the Th1 response mediating graft rejection. In addition, these pathways may function in promoting alloantibody responses, which could contribute to inflammatory responses in vessels.

THE PD-1/PD-L PATHWAY PD-1 (program death-1) is a 50–55 kDa type I transmembrane receptor that was identified in a T cell line undergoing activation-induced cell death (62). PD-1 is a member of the Ig superfamily that contains a single Ig V–like domain in

4 Feb 2002

11:14

AR

AR152-02.tex

AR152-02.SGM

LaTeX2e(2001/05/10)

P1: GSR

Annu. Rev. Immunol. 2002.20:29-53. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

NEW B7-LIKE LIGANDS AND RECEPTORS

39

its extracellular region (Table 1); it lacks the MYPPPY motif, a sequence critical for CTLA-4 and CD28 binding to B7.1 and B7.2 (63). The PD-1 cytoplasmic domain contains two tyrosines, with the most membrane-proximal tyrosine (VAYEEL in mouse PD-1) located within an ITIM (immuno-receptor tyrosinebased inhibitory motif) (62). The presence of an ITIM on PD-1 strongly suggested that this molecule could function to attenuate antigen receptor signaling by recruitment of cytoplasmic phosphatases (64). Human and murine PD-1 proteins share 60% amino acid identity with conservation of four potential N-glycosylation sites, and residues that define the Ig-V domain (65, 66). The ITIM in the cytoplasmic region and the ITIM-like motif surrounding the carboxy-terminal tyrosine (TEYATI in human and mouse) are also conserved between human and murine orthologues. There is 62% amino acid identity in the PD-1 cytoplasmic region between human and murine proteins. This contrasts with the 100% conservation observed between human and murine CTLA-4 cytoplasmic regions. The genome location of human PD-1 has been mapped to chromosome 2q37.3 (65); CTLA-4, CD28, and ICOS mapped on the same chromosome at 2q33 (17, 22, 22a). In normal murine tissue, PD-1 mRNA expression is confined to the thymus (67). Approximately 1% of thymocytes are PD-1 positive with expression restricted to a subset within the double negative (DN) population (67). In vivo administration of anti-CD3 mAb results in marked apoptosis of CD4+CD8+ (DP) cells in the thymus; however, such treatment leads to induction of PD-1 protein expression on the surviving CD4−CD8− (DN) and single positive (SP) thymocytes. Significantly anti-CD3 mAb, but not dexamethasome, treatment induces PD-1 expression. These observations strongly suggest that T cell activation, and not induction of apoptotic death per se, results in expression of PD-1. The potential role of PD-1 in thymic selection has been studied in detail (68). An increased percentage of PD-1+ DN thymocytes has been reported in neutral and positively selecting backgrounds. Interestingly, PD-1 deficiency in positive selecting backgrounds resulted in an increase in DP cells and a decrease in SP thymocytes (68). These findings suggested a role for PD-1 in thymic positive selection, in which PD-1 engagement could increase the threshold of pre-TCR/CD3 complex signals required for transition from the DN to DP stage. In PD-1-deficient mice, a lower pre-TCR/CD3 complex threshold would allow for a higher number of cells to transition from DN to DP. In addition, PD-1 could also affect the efficiency of positive selection by modulating the threshold of TCR αβ signals. In contrast, PD-1 deficiency has a negligible effect on negative selection of TCR transgenic T cells (69). These studies suggest that TCR thresholds of activation can be modulated upon engagement of PD-1. Moreover, these studies suggest that PD-1 plays no significant role in central tolerance. Analysis of murine spleen and lymph node populations indicated that within each of these populations a small percentage of PD-1 positive cells could be detected (63). Under resting conditions, neither T nor B cells expressed PD-1. However, activation of T or B cells through the antigen receptor or with PMA and

4 Feb 2002

11:14

Annu. Rev. Immunol. 2002.20:29-53. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

40

AR

AR152-02.tex

CARRENO

¥

AR152-02.SGM

LaTeX2e(2001/05/10)

P1: GSR

COLLINS

ionomycin resulted in cell surface expression of the PD-1 receptor (63). PD-1+ T and B cells are large in size and co-express CD25 and CD69, correlating expression of PD-1 with cell activation. PD-1 protein expression can be detected as early as 24 h after TCR activation. Both CD4+ and CD8+ T cell populations expressed cell surface PD-1 upon activation, and the pattern of PD-1 expression on human T cells is similar to that observed for murine T cells (B. Carreno, unpublished observations). PD-1 is also expressed in activated macrophages (70). The ligands for PD-1 are the B7 family members PD-L1 (B7-H1) and PD-L2 (B7-DC) (71–74). Interaction of PD-1 with either PD-L1 or PD-L2 results in inhibition of T and B cell responses (70–72). An alternatively spliced PD-L2 variant lacking the Ig V–like domain has been described, but this variant does not bind to PD-1 (72). PD-L1 and PD-L2 share 40% amino acid identity and thus are more homologous to each other than to other ligands of the B7 family (Table 2). Human and murine orthologues of PD-L1 or PD-L2 share 70% amino acid identity. Interestingly, murine PD-L2 has only 5 amino acids in its cytoplasmic tail, whereas the human PD-L2 cytoplasmic region is 28 amino acids in length. Both human PD-L1 and PD-L2 genes map to chromosome 9p24.2, and these genes are separated by only 42 kb (72). This genomic proximity is reminiscent of that observed for B7-1 and B7-2, which are tightly linked on human chromosome 3q13.3-21 (41). Murine PD-L2 maps to a region located between 19C2 and 19C3 (74). A variety of normal tissues have been examined for expression of PD-L1 and PD-L2 transcripts (71–73). The pattern of expression of these molecules is significantly broader than that reported for other B7 family ligands. The overall distribution of PD-L1 and PD-L2 transcripts is similar in human and murine tissues, with high levels of expression in placenta, low expression levels in spleen, lymph nodes, and thymus, and the absence of expression in brain. Transcripts for both PD-L1 and PD-L2 are detected in human heart; in murine hearts, transcripts for PD-L1 are abundantly expressed whereas PD-L2 transcripts are absent (71, 72). PD-L2 but not PD-L1 transcripts are detected in human pancreas, lung, and liver (72). Identification of the cellular populations that express PD-L1 and PD-L2 in these tissues awaits further investigation. Expression of PD-L1 and PD-L2 in both lymphoid and nonlymphoid tissues suggests that the PD-1/PD-L pathway may modulate immune responses in secondary lymphoid organs as well as in peripheral sites. Expression of PD-L1 and PD-L2 on antigen presenting cell (APC) populations has also been examined in detail (71–73). Resting B cells, monocytes, and dendritic cells do not express either PD-L1 or PD-L2. Transcripts for these ligands can be detected upon activation of these populations by antigen receptor, LPS, or IFN-γ . In human B cell populations, LPS or BCR activation results in induction of PD-L1 and PD-L2 (71–73). In human monocytes, IFN-γ , but not TNF-α, treatment results in expression of both ligands; PD-L1 expression precedes that of PD-L2. Interestingly, IFN-γ treatment also upregulates B7-1 transcripts and ICOS-L protein expression on human monocyte populations (46, 71). On dendritic cells, LPS plus IFN-γ treatment induces PD-L1 and PD-L2 mRNA expression (71–73). B7-1 and B7-2 transcripts are also upregulated in these cells by LPS plus IFN-γ . Tseng

4 Feb 2002

11:14

AR

AR152-02.tex

AR152-02.SGM

LaTeX2e(2001/05/10)

P1: GSR

Annu. Rev. Immunol. 2002.20:29-53. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

NEW B7-LIKE LIGANDS AND RECEPTORS

41

et al. (74) reported that PD-L2 (B7-DC) transcripts are exclusively expressed on spleen and bone marrow–derived murine dendritic cells. These authors postulate that some of the unique functions of dendritic cells could be attributed to PD-L2 expression. Further studies will be necessary to sort out these discrepancies in expression patterns. Interestingly, mitogen or TCR activation of murine and human T cells results in cell surface expression of PD-L1 in addition to PD-1 (73) (L. Carter, B. Carreno, unpublished observations). Similarly, BCR-activated B cells express PD-L1 and PD-L2 in addition to PD-1 (63, 71, 72). Thus, B and T cell function can be modulated by engagement of cell-surface PD-1. Additionally, upon activation, both T and B cells can engage PD-1 on other cells through expression of PD-1 ligands. This suggests that at points of T:B contact, modulation of antigen receptor signals can occur bidirectionally through PD-1. The consequences of such an interaction are unknown but might serve to limit TCR and BCR receptor signaling after activation. IFN-γ can also modulate PD-L1 expression in nonlymphoid cells. Endothelial cells constituitively express cell-surface PD-L1, and in vitro treatment with IFN-γ , but not LPS or TNF-α, results in its rapid upregulation (M. Eppihimer, J. Leonard, personal communication). Furthermore, IL-12 challenge of IFN-γ +/+ but not IFN-γ −/− mice results in enhanced expression of PD-L1 in blood vessels of various tissues (M. Eppihimer, J. Leonard, personal communication). Thus in vivo, IFN-γ upregulation of PD-L1 expression on endothelial cells may play a significant role in attenuation of lymphocyte function at peripheral sites. Consistent with these findings, several studies have suggested an immunosuppressive role for IFN-γ . IFN-γ receptor–deficient mice develop accelerated collagen-induced arthritis (75), and IFN-γ blockade enhances EAE (76). In addition, IFN-γ treatment has been reported to confer resistance to EAE (77). Thus, the beneficial effect of IFN-γ reported in some autoimmune settings could be partly attributed to induction of PD-L1 ligand expression at sites of inflammation and subsequent downregulation of immune responses by PD-1 engagement. Finally, PD-L1 and PD-L2 transcripts have been detected in various tumor cell lines (72). Additionally, cell-surface expression of PD-L1 has been reported in human breast cancer cell lines (72). These observations have led to the suggestion that tumors may escape immunosurveillance by attenuation of T cell responses upon PD-1 engagement. This hypothesis has implications for the development of new strategies for tumor immunotherapy, as one would predict that blockade of PD-1/PD-L interactions could enhance tumor-specific T cell responses.

ATTENUATION OF IMMUNE RESPONSES BY THE PD-1/PD-L PATHWAY Identification of the PD-1 ligands, PD-L1 (B7-H1) and PD-L2 (B7-DC), and assessment of their interaction with PD-1 confirmed the negative regulatory function of PD-1 in immune responses (71, 72). Neither PD-L1 nor PD-L2 bound to CD28,

4 Feb 2002

11:14

Annu. Rev. Immunol. 2002.20:29-53. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

42

AR

AR152-02.tex

CARRENO

¥

AR152-02.SGM

LaTeX2e(2001/05/10)

P1: GSR

COLLINS

CTLA-4, or ICOS. Reciprocally, soluble forms of B7-1 and B7-2 did not bind PD-1. Experiments using T cells from PD-1-deficient mice formally demonstrated the functional consequences of PD-1/PD-L1 interactions (71). Proliferation of wild-type, but not PD-1-deficient, T cells was inhibited in the presence of PD-L1. Similarly, proliferation of human T cells was decreased in the presence of PD-L1. Cross-linking of PD-1 by PD-L2 upon TCR activation also resulted in decreased proliferation (72). Furthermore, PD-1 cross-linking by either ligand resulted in decreased IFN-γ , IL-10, IL-4, and IL-2 secretion (71, 72). These results indicate that, upon TCR activation, cross-linking of PD-1 by PD-L1 or PD-L2 leads to diminished immune responses. Thus, PD-1 cross-linking by either PD-L1 or PD-L2 results in similar outcomes, suggesting that these two ligands may have overlapping functions in vivo. However, additional information regarding their affinities for PD-1, as well as the generation of PD-L1- and PD-L2-deficient mice will be necessary to discern their potential roles in vivo. Consistent with an inhibitory function, studies on PD-1+ B cell lymphomas have shown that PD-1/BCR co-engagement results in inhibition of Ca2+ influx as well as hypophosphorylation of BCR downstream signaling molecules syk, phosphatidyl inositol-3, phospholipase C, and vav (70). SHP-1 and SHP-2, Src-homology-2 (SH-2) domain containing phosphatases, have been implicated in inhibitory signals mediated by NK receptors, with recruitment of both phosphatases upon tyrosine phosphorylation of ITIM motifs (64). A role for SHP-2 in CTLA-4 inhibitory signals has also been reported (78). Co-ligation of the BCR and PD-1 resulted in increased phosphorylation of SHP-2 and recruitment of SHP-2 to the PD-1 receptor (70). Interestingly, the carboxy-terminal tyrosine (TEYATI), and not the tyrosine within the canonical ITIM, has been implicated as necessary for SHP-2 binding (70). Similarly, TCR/PD-1 co-ligation resulted in increased phosphorylation of SHP-2 (72). Thus, these studies point to SHP-2 as a likely candidate involved in transducing inhibitory signals initiated by PD-1. PD-1 appears to function as an attenuator of T cell responses, and the process by which this regulation occurs is distinct from activation-induced cell death (72). PD-1 engagement by either PD-L1 or PD-L2 results in cell cycle arrest. Activation of cells in the presence of either ligand leads to an accumulation of cells at the G0/G1 phase of the cell cycle. IL-2 production is drastically inhibited upon PD-1 engagement, while exogenous IL-2 can rescue PD-1-mediated cell cycle arrest (L. Carter, B. Carreno, unpublished observations). These observations suggest that PD-1 may affect T cell activation and proliferation by regulating IL-2 transcription. These data parallel those reported for CTLA-4, which most likely inhibits T cell cycle entry by regulating IL-2 transcription and mRNA stability (79–81). The interplay between PD-1 and CD28 on T cell activation has also been examined. Optimal, but not suboptimal, CD28 costimulation can rescue PD-1-mediated inhibition (71). At low antigen concentrations, PD-1 signals can antagonize costimulation mediated by CD28. At high antigen levels, CD28 costimulation overrides the PD-1 inhibitory effect. Interestingly, PD-1 cell surface expression is highest

4 Feb 2002

11:14

AR

AR152-02.tex

AR152-02.SGM

LaTeX2e(2001/05/10)

P1: GSR

Annu. Rev. Immunol. 2002.20:29-53. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

NEW B7-LIKE LIGANDS AND RECEPTORS

43

at low antigen concentration. Thus, a correlation can be established between antigen concentrations required for highest PD-1 expression and conditions in which downregulation of T cell responses are more readily observed (72). Interestingly, two separate groups have concluded that PD-L1 (B7-H1) and PD-L2 (BC-DC) can function to costimulate T cell responses (73, 74). Both groups have reported that activation of T cells in the presence of suboptimal TCR signals, and either PD-L1 or PD-L2 results in increased proliferation. Additionally, anti-CD3 plus PD-L1.Fc activation results in increased secretion of IL-10, IFN-γ and GM-CSF but not IL-2 or IL-4 (73). Activation of T cells with anti-CD3 and PD-L2.Fc increases IFN-γ but not IL-4 or IL-10 (74). The discrepancy between these results and those of Freeman and colleagues (71, 72) raises the possibility that there may be additional receptors for PD-L1 and PD-L2. If, indeed, a second receptor with costimulatory function exists for PD-L ligands, this pathway would have symmetry with that of CD28/CTLA-4/B7. As speculated for CD28/CTLA-4, temporal regulation of receptor expression and affinities of ligands would then determine whether the costimulatory or inhibitory signals prevail.

PD-1 DEFICIENCY LEADS TO AUTOIMMUNE DISORDERS AND BREAKDOWN OF PERIPHERAL TOLERANCE Consistent with its negative regulatory function, PD-1 deficiency in vivo results in the development of autoimmune disorders (69, 82, 83). C57BL/6- PD-1−/− mice consistently displayed splenomegaly, increased numbers of B lymphocytes and myeloid cells, and increased serum IgG2b, IgG3, and IgA (82). Antibody responses to T-independent, but not T-dependent, antigens were greatly enhanced in PD-1−/− mice relative to control littermates. Furthermore, PD-1−/− B cells displayed enhanced proliferation in response to BCR cross-linking. These mice spontaneously developed a lupus-like disease with age (69). At 6 months, PD-1-deficient mice displayed elevated serum IgG3 levels and increased IgG3 and C3 deposition in the glomeruli. At 14 months, approximately 50% of these mice had lupus-like glomerulonephritis and histological evidence of arthritis as well as granulomatous inflammation. Introduction of the lpr mutation (B6-lpr/lpr-PD-1−/− ) accelerated the onset and severity of disease. Thus, PD-1 deficiency in the C57BL/6 background resulted in the development of a late onset, chronic, progressive, lupuslike glomerulonephritis and arthritis, and the severity of disease was exacerbated by the absence of FAS-mediated apoptosis. Interestingly, introduction of the PD-1 deficiency into the Balb/c background resulted in a distinct autoimmune phenotype, with cardiomegaly, diffuse IgG1 deposition in cardiomyocytes, and high circulating levels of heart-tissue reactive IgG1 (83). This disease developed rapidly, and Balb/c-PD-1−/− mice died as early as 5 weeks of age. By 30 weeks, two thirds of mice had succumbed to disease. In contrast, no disease was observed in Balb/c-PD-1−/− -RAG−/− mice, indicating that T and B cells are required for disease development. The distinct severity and phenotype of disease observed in

4 Feb 2002

11:14

Annu. Rev. Immunol. 2002.20:29-53. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

44

AR

AR152-02.tex

CARRENO

¥

AR152-02.SGM

LaTeX2e(2001/05/10)

P1: GSR

COLLINS

these two strains indicated that other genetic modifier genes in addition to the PD-1 deficiency contribute to the pathologies observed. However, in both genetic backgrounds, PD-1 deficiency appeared to alter the balance between activating and inhibitory signals, resulting in a loss of peripheral tolerance. Further support for a role for PD-1 in the maintenance and/or induction of peripheral tolerance has come from studies examining responses to self-antigens (69). PD-1 deficient 2C TCR transgenic mice bred to the autoreactive background (H-2b/d) exhibited growth retardation, splenomegaly, and lethal graft-vs-host disease. Massive infiltration of inflammatory cells in liver, heart, and lung was observed. Additionally, an increase in the total number of cells, most notably activated CD8+ T cells, in the spleens was also reported. As PD-1 has a minimal, if any, role in negative selection in the thymus (69), these results point to PD-1 as negative regulator of self-reactivity in the periphery. Altogether, the outcome is a breakdown of peripheral tolerance to selected tissue antigens. In vitro, PD-1 and CTLA-4 functions are quite similar. Engagement of either PD-1 or CTLA-4 results in inhibition of T cell proliferation, cytokine production, and cell cycle progression (3, 71). In vivo, deficiency in these molecules results in development of lymphoproliferative disorders, albeit with different degrees of severity. CTLA-4-deficient mice display very aggressive lymphoproliferative disorders and die at 21–28 days of age (6, 7, 84). Lymphocytic infiltration is observed in multiple organs. The disorder is characterized by a high frequency of T cell blasts (CD25, CD69, CD44hi, CD45ROlow), with signs of T cell activation detected as early as 5–6 days after birth. In contrast, a less aggressive disorder is observed in PD-1-deficient mice (69, 83). Of the phenotypes reported, the most aggressive is the cardiomyopathy in Balb/c-PD-1−/− mice (83). As deficiencies in either of these receptors result in breakdown of peripheral tolerance, both receptors must have critical and nonredundant functions in the maintenance of tolerance. These pathways may control T cell responses at two distinct points, with the first at the time of T cell activation in lymphoid tissues, and the second upon reactivation in peripheral sites. CTLA-4 would have a predominant role in regulating the threshold for T cell activation in lymphoid sites, where B7-1 and B7-2 are primarily expressed. Because both PD-1 and PD-1 ligands have broader expression patterns, PD-1 could be important in regulating thresholds of activation for T and B cells in both lymphoid and peripheral sites during inflammation. By controlling the magnitude of T cell responses at initiation and again at reactivation, these pathways could function as independent checkpoints to safeguard against self-reactivity.

B7-H3: A NEW B7-LIKE LIGAND Human B7-H3 was recently identified as a new costimulatory member of the B7 family (85). B7-H3 shares from 26% to 33% amino acid identity with other members of the B7 family (Table 2). Northern analysis indicated that this gene is broadly expressed, with mRNA detected in most organs, as well as in immune tissues

4 Feb 2002

11:14

AR

AR152-02.tex

AR152-02.SGM

LaTeX2e(2001/05/10)

P1: GSR

Annu. Rev. Immunol. 2002.20:29-53. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

NEW B7-LIKE LIGANDS AND RECEPTORS

45

including spleen, lymph node, thymus, bone marrow, and fetal liver. Tumor cell lines derived from nonlymphoid tissues were also positive for B7-H3 mRNA, but Molt-4, a lymphoblastic leukemia, and Raji, a Burkitt’s lymphoma line, were not. B7-H3 expression was not detected in unstimulated PBL but could be induced in lymphoid cells by activation. B7-H3 protein was detected on GM-CSF-stimulated monocytes and IFN-γ activated dendritic cells, as well as on CD3+ T cells activated with PMA and ionomycin. B7-H3-transfected 293 cells did not bind CTLA-4.Fc, ICOS.Fc, or PD-1.Fc, which suggests that this B7 member binds a distinct receptor (85). Binding studies with the B7-H3.Fc protein further suggest that this receptor is present on activated T cells. Activation of human T cells with plate-bound anti-CD3 plus increasing concentrations of B7-H3.Fc resulted in a dose-dependent enhancement of proliferation relative to cells activated with anti-CD3 plus control Ig. B7-H3 enhancement of proliferative responses was not as profound as that observed with B7-1.Fc. B7-H3.Fc enhances proliferative responses of both CD4+ and CD8+ T cells. In addition, B7-H3.Fc stimulation increased secretion of IFNγ in 50% of cell donors tested. Interestingly, expression of B7-H3 in a melanoma line resulted in increased lytic activity of melanoma-specific T cells. The receptor for this newest member of the B7 family remains to be identified, and further elucidation of B7-H3 function in immune responses is needed.

NEW MEMBERS OF THE B7 AND CD28 FAMILY? The recent expansion of the B7 family and its receptors raises the question of whether there are additional members of this family with immune function. Certainly, the receptor for B7-H3 remains to be identified. Conflicting results indicating that PD-L1 and PD-L2 can both costimulate and inhibit immune responses raise the possibility of additional receptors for these ligands. In addition, experiments demonstrating B7-dependent responses in mice deficient for CD28 and CTLA-4 suggest that there are additional receptors for the B7 molecules (86, 87). Finally, experiments evaluating ICOS-L often use the ICOS.Fc reagent to define this ligand. Antibody reagents for ICOS-L are needed to confirm that these interactions do not include contributions of additional ligands for ICOS. In addition to the B7-like ligands and their receptors described above, molecules with homologies to these families have been described. In each of these cases, data are too limited to currently include these molecules as bona fide members of the B7-family and their receptors. A key question is whether these related proteins have a role in immune responses. Members of the butyrophilin gene family share homology with the B7-family of immunoregulatory receptors (88, 89). Butyrophilin is a 66-kDa type I transmembrane protein that forms a major component of milk fat globule membrane (90, 91). It has no reported role in the immune system. Six additional family members were identifed by homology and genomic location and form two subfamilies (92, 93). The seven members of the butyrophilin gene family are located in a cluster

4 Feb 2002

11:14

Annu. Rev. Immunol. 2002.20:29-53. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

46

AR

AR152-02.tex

CARRENO

¥

AR152-02.SGM

LaTeX2e(2001/05/10)

P1: GSR

COLLINS

on human chromosome 6p22.1, telomeric to the class I MHC (89, 93). Like the B7 proteins, butyrophilins are type I transmembrane proteins with IgV and IgC extracellular domains. However, they also have unique cytoplasmic domains, including a series of heptad repeats and a B30.2 domain, which may mediate protein interactions (89, 93). Butyrophilin genes are expressed at low levels in most tissues (93). Interestingly, the predominant transcripts of the BTN3 subfamily do not express the B30.2 domain due to alternate splicing (93), making the BTN3 subfamily proteins more similar to the B7 gene family. Linsley et al. (88) also noted that the myelin protein MOG has homology with B7-1, B7-2, and butyrophilin, although MOG contains only an IgV extracellular domain. MOG is an autoantigen in EAE, and immunization of susceptible rats with the IgV domain of butyrophilin also results in an inflammatory response in the CNS due to cross-reactivity of T cells to MOG-derived epitopes (94). MOG maps to 6p21 in humans, centromeric to the butyrophilin genes and telomeric from the class I and II gene clusters (95, 96). Although MOG can act as an autoantigen in animal models of multiple sclerosis, as yet there are no data defining an interacting protein for MOG, nor any data implicating MOG as a costimulatory protein in immune function. Beyond these defined family members, other potential homologous sequences have been identified in database searches (97). These include SIRPα and β, transmembrane proteins expressed in the immune system; HHLA2, an endogenous retroviral sequence; MCAM, a melanoma adhesion protein; and VEJAM, vascular endothelial junction–associated molecule, as well as a few novel sequences with similar levels of homology. Additionally, Linsley et al. (88) noted the homology of the chicken B-G gene to the B7 family. Homology searches to the CD28 receptor family can also be carried out, and there is sufficient homology between CD28, ICOS, and CTLA-4 to detect these as related sequences. However, PD-1 is less homologous to this family, suggesting that the receptors for the B7 family of ligands may be less conserved. Indeed, as each of these receptors contains only a single extracellular IgV domain, many proteins of the Ig-superfamily can be detected with weak homology, including TCRα and Igκ proteins. This suggests that functional tests of members of the Ig superfamily will be critical in identifying new members of costimulatory and inhibitory receptors and ligands related to CD28 and B7.

PERSPECTIVES The sequencing of the human genome has led to an explosion of gene identification, along with information about gene clusters. The new challenge is to ascribe function to new genes with sequence relationships to known genes. For the family of genes related to the CD28 and B7 molecules, this has led to an appreciation that these receptor-ligand interactions will include both costimulatory and inhibitory receptors, and that multiple receptor-ligand interactions are possible. In addition, regulation of the immune response can be affected by signals delivered through

4 Feb 2002

11:14

AR

AR152-02.tex

AR152-02.SGM

LaTeX2e(2001/05/10)

P1: GSR

NEW B7-LIKE LIGANDS AND RECEPTORS

47

Annu. Rev. Immunol. 2002.20:29-53. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

multiple costimulatory or inhibitory receptors, each of which is simultaneously expressed by the interacting cells. Integration of these various signals with signals through antigen-specific receptors then determines the outcome of that cellular interaction. Regulation of immune responses in both the priming events in immune sites and in activation or attenuation in peripheral sites can occur through these pathways. An understanding of which interactions are critical at various steps in immune responses will allow intervention in immune-mediated diseases through precise manipulation of these pathways. ACKNOWLEDGMENTS We would like to thank Dr. Vincent Ling, Dr. Laura Carter, and Dr. Arlene Sharpe for critical reading of the manuscript and for many helpful discussions. We would also like to thank our colleagues at Genetics Institute/Wyeth Research and our academic collaborators for many enjoyable and thought-provoking discussions in the field of costimulation and immune responses. Visit the Annual Reviews home page at www.annualreviews.org

LITERATURE CITED 1. Salomon B, Bluestone JA. 2001. Complexities of CD28/B7: CTLA-4 costimulatory pathways in autoimmunity and transplantation. Annu. Rev. Immunol. 19: 225–52 2. Greenfield EA, Nguyen KA, Kuchroo VK. 1998. CD28/B7 costimulation: a review. Crit. Rev. Immunol. 18:389–418 3. Chambers CA, Kuhns MS, Egen JG, Allison JP. 2001. CTLA-4-mediated inhibition in regulation of T cell responses: mechanisms and manipulation in tumor immunotherapy. Annu. Rev. Immunol. 19: 565–94 4. Boise LH, Minn AJ, Noel PJ, June CH, Accavitti MA, Lindsten T, Thompson CB. 1995. CD28 costimulation can promote T cell survival by enhancing the expression of Bcl-XL. Immunity 3:87– 98 5. Linsley PS, Bradshaw J, Greene J, Peach R, Bennett KL, Mittler RS. 1996. Intracellular trafficking of CTLA-4 and focal localization towards sites of TCR engagement. Immunity 4:535–43

6. Tivol EA, Borriello F, Schweitzer AN, Lynch WP, Bluestone JA, Sharpe AH. 1995. Loss of CTLA-4 leads to massive lymphoproliferation and fatal multiorgan tissue destruction, revealing a critical negative regulatory role of CTLA-4. Immunity 3:541–47 7. Waterhouse P, Penninger JM, Timms E, Wakeham A, Shahinian A, Lee KP, Thompson CB, Griesser H, Mak TW. 1995. Lymphoproliferative disorders with early lethality in mice deficient in CTLA4. Science 270:985–88 8. Chambers CA, Sullivan TJ, Allison JP. 1997. Lymphoproliferation in CTLA-4deficient mice is mediated by costimulation-dependent activation of CD4+ T cells. Immunity 7:885–95 9. van der Merwe PA, Bodian DL, Daenke S, Linsley P, Davis SJ. 1997. CD80 (B7-1) binds both CD28 and CTLA-4 with a low affinity and very fast kinetics. J. Exp. Med. 185:393–403 10. Ikemizu S, Gilbert RJ, Fennelly JA, Collins AV, Harlos K, Jones EY, Stuart

4 Feb 2002

11:14

48

11.

Annu. Rev. Immunol. 2002.20:29-53. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

12.

13.

14.

15.

16.

17.

18.

AR

AR152-02.tex

CARRENO

¥

AR152-02.SGM

LaTeX2e(2001/05/10)

P1: GSR

COLLINS

DI, Davis SJ. 2000. Structure and dimerization of a soluble form of B7-1. Immunity 12:51–60 Borriello F, Sethna MP, Boyd SD, Schweitzer AN, Tivol EA, Jacoby D, Strom TB, Simpson EM, Freeman GJ, Sharpe AH. 1997. B7-1 and B7-2 have overlapping, critical roles in immunoglobulin class switching and germinal center formation. Immunity 6:303–13 Schwartz JC, Zhang X, Fedorov AA, Nathenson SG, Almo SC. 2001. Structural basis for co-stimulation by the human CTLA-4/B7-2 complex. Nature 410:604–8 Stamper CC, Zhang Y, Tobin JF, Erbe DV, Ikemizu S, Davis SJ, Stahl ML, Seehra J, Somers WS, Mosyak L. 2001. Crystal structure of the B7-1/CTLA-4 complex that inhibits human immune responses. Nature 410:608–11 Hutloff A, Dittrich AM, Beier KC, Eljaschewitsch B, Kraft R, Anagnostopoulos I, Kroczek RA. 1999. ICOS is an inducible T-cell co-stimulator structurally and functionally related to CD28. Nature 397:263–66 Tamatani T, Tezuka K, Hanzawa-Higuchi N. 2000. AILIM/ICOS: a novel lymphocyte adhesion molecule. Int. Immunol. 12: 51–55 Beier KC, Hutloff A, Dittrich AM, Heuck C, Rauch A, Buchner K, Ludewig B, Ochs HD, Mages HW, Kroczek RA. 2000. Induction, binding specificity and function of human ICOS. Eur. J. Immunol. 30:3707–17 Mages HW, Hutloff A, Heuck C, Buchner K, Himmelbauer H, Oliveri F, Kroczek RA. 2000. Molecular cloning and characterization of murine ICOS and identification of B7h as ICOS ligand. Eur. J. Immunol. 30:1040–47 Tezuka K, Tsuji T, Hirano D, Tamatani T, Sakamaki K, Kobayashi Y, Kamada M. 2000. Identification and characterization of rat AILIM/ICOS, a novel Tcell costimulatory molecule, related to the

19.

20.

21.

22.

22a.

23.

24.

CD28/CTLA4 family. Biochem. Biophys. Res. Commun. 276:335–45 Bajorath J, Metzler WJ, Linsley PS. 1997. Molecular modeling of CD28 and three-dimensional analysis of residue conservation in the CD28/CD152 family. J. Mol. Graph. Model. 15:135–39, 108–11 Ling V, Wu PW, Finnerty HF, Bean KM, Spaulding V, Fouser LA, Leonard JP, Hunter SE, Zollner R, Thomas JL, Miyashiro JS, Jacobs KA, Collins M. 2000. Cutting edge: identification of GL50, a novel B7-like protein that functionally binds to ICOS receptor. J. Immunol. 164:1653–57 Yoshinaga SK, Whoriskey JS, Khare SD, Sarmiento U, Guo J, Horan T Shih G, Zhang M, Coccia MA, Kohno T, TafuriBladt A, Brankow D, Campbell P, Chang D, Chiu L, Dai T, Duncan G, Elliott GS, Hui A, McCabe SM, Scully S, Shahinian A, Shaklee CL, Van G, Mak TW, Senaldi G. 1999. T-cell co-stimulation through B7RP-1 and ICOS. Nature 402:827–32 Coyle AJ, Lehar S, Lloyd C, Tian J, Delaney T, Manning S, Nguyen T, Burwell T, Schneider H, Gonzalo JA, Gosselin M, Owen LR, Rudd CE, Gutierrez-Ramos JC. 2000. The CD28-related molecule ICOS is required for effective T celldependent immune responses. Immunity 13:95–105 Ling V, Wu PW, Finnerty HF, Agostino MJ, Graham JR, Chen S, Fisk GJ, Miller CP, Collins M. 2001. Assembly and annotation of human chromosome 2q33 sequence containing the CD28, CTLA4, and ICOS gene cluster: analysis by computational, comparative and microarray approaches. Genomics. In press Agarwal S, Rao A. 1998. Modulation of chromatin structure regulates cytokine gene expression during T cell differentiation. Immunity 9:765–75 Buonfiglio D, Bragardo M, Redoglia V, Vaschetto R, Bottarel F, Bonissoni S, Bensi T, Mezzatesta C, Janeway CA Jr, Dianzani U. 2000. The T cell activation

4 Feb 2002

11:14

AR

AR152-02.tex

AR152-02.SGM

LaTeX2e(2001/05/10)

P1: GSR

NEW B7-LIKE LIGANDS AND RECEPTORS

Annu. Rev. Immunol. 2002.20:29-53. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

25.

26.

27.

28.

29.

30.

31.

molecule H4 and the CD28-like molecule ICOS are identical. Eur. J. Immunol. 30: 3463–67 McAdam AJ, Chang TT, Lumelsky AE, Greenfield EA, Boussiotis VA, DukeCohan JS, Chernova T, Malenkovich N, Jabs C, Kuchroo VK, Ling V, Collins M, Sharpe AH, Freeman GJ. 2000. Mouse inducible costimulatory molecule (ICOS) expression is enhanced by CD28 costimulation and regulates differentiation of CD4+ T cells. J. Immunol. 165:5035–40 Kopf M, Coyle AJ, Schmitz N, Barner M, Oxenius A, Gallimore A, GutierrezRamos JC, Bachmann MF. 2000. Inducible costimulator protein (ICOS) controls T helper cell subset polarization after virus and parasite infection. J. Exp. Med. 192:53–61 Okkenhaug K, Wu L, Garza KM, La Rose J, Khoo W, Odermatt B, Mak TW, Ohashi PS, Rottapel R. 2001. A point mutation in CD28 distinguishes proliferative signals from survival signals. Nat. Immunol. 2:325–32 Burr JS, Savage ND, Messah GE, Kimzey SL, Shaw AS, Arch RH, Green JM. 2001. Cutting edge: distinct motifs within CD28 regulate T cell proliferation and induction of Bcl-X(l). J. Immunol. 166:5331–35 Holdorf AD, Green JM, Levin SD, Denny MF, Straus DB, Link V, Changelian PS, Allen PM, Shaw AS. 1999. Proline residues in CD28 and the Src homology (SH)3 domain of Lck are required for T cell costimulation. J. Exp. Med. 190: 375–84 Breitfeld D, Ohl L, Kremmer E, Ellwart J, Sallusto F, Lipp M, Forster R. 2000. Follicular B helper T cells express CXC chemokine receptor 5, localize to B cell follicles, and support immunoglobulin production. J. Exp. Med. 192:1545–52 Schaerli P, Willimann K, Lang AB, Lipp M, Loetscher P, Moser B. 2000. CXC chemokine receptor 5 expression defines follicular homing T cells with B cell

32.

33.

34.

35.

36.

37.

38.

39.

49

helper function. J. Exp. Med. 192:1553– 62 McAdam AJ, Greenwald RJ, Levin MA, Chernova T, Malenkovich N, Ling V, Freeman GJ, Sharpe AH. 2001. ICOS is critical for CD40-mediated antibody class switching. Nature 409:102–4 Dong C, Juedes AE, Teman UA, Shresta S, Allison JP, Ruddle NH, Flavell RA. 2001. ICOS co-stimulatory receptor is essential for T cell activation and function. Nature 409:97–101 Tafuri A, Shahinian A, Bladt F, Yoshinaga SK, Jordana M, Wakeham A, Boucher LM, Bouchard D, Chan VSF, Duncan G, Odermatt B, Ho A, Itie A, Horan T, Whoriskey JS, Pawson T, Penninger JM, Ohashi PS, Mak TW. 2001. ICOS is essential for effective T-helper-cell responses. Nature 409:105–8 Shahinian A, Pfeffer K, Lee KP, Kundig TM, Kishihara K, Wakeham A, Kawai K, Ohashi PS, Thompson CB, Mak TW. 1993. Differential T cell costimulatory requirements in CD28-deficient mice. Science 261:609–12 Swallow MM, Wallin JJ, Sha WC. 1999. B7h, a novel costimulatory homolog of B7.1 and B7.2, is induced by TNFalpha. Immunity 11:423–32 Wang S, Zhu G, Chapoval AI, Dong H, Tamada K, Ni J, Chen L. 2000. Costimulation of T cells by B7-H2, a B7-like molecule that binds ICOS. Blood 96: 2808–13 Brodie D, Collins AV, Iaboni A, Fennelly JA, Sparks LM, Xu XN, van der Merwe PA, Davis SJ. 2000. LICOS, a primordial costimulatory ligand? Curr. Biol. 10:333–36 Ishikawa K, Nagase T, Suyama M, Miyajima N, Tanaka A, Kotani H, Nomura N, Ohara O. 1998. Prediction of the coding sequences of unidentified human genes. X. The complete sequences of 100 new cDNA clones from brain which can code for large proteins in vitro. DNA Res. 5: 169–76

4 Feb 2002

11:14

Annu. Rev. Immunol. 2002.20:29-53. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

50

AR

AR152-02.tex

CARRENO

¥

AR152-02.SGM

LaTeX2e(2001/05/10)

P1: GSR

COLLINS

40. Hattori M, Fujiyama A, Taylor TD, Watanabe H, Yada T, Park HS, Toyoda A, Ishii K, Totoki Y, Choi DK, Soeda E, Ohki M, Takagi T, Sakaki Y, Taudien S, Blechschmidt K, Polley A, Menzel U, Delabar J, Kumpf K, Lehmann R, Patterson D, Reichwald K, Rump A, Schillhabel M, Schudy A, Zimmermann W, Rosenthal A, Kudoh J, Schibuya K, Kawasaki K, Asakawa S, Shintani A, Sasaki T, Nagamine K, Mitsuyama S, Antonarakis SE, Minoshima S, Shimizu N, Nordsiek G, Hornischer K, Brant P, Scharfe M, Schon O, Desario A, Reichelt J, Kauer G, Blocker H, Ramser J, Beck A, Klages S, Hennig S, Riesselmann L, Dagand E, Haaf T, Wehrmeyer S, Borzym K, Gardiner K, Nizetic D, Francis F, Lehrach H, Reinhardt R, Yaspo ML. 2000. The DNA sequence of human chromosome 21. Nature 405:311–19 41. Reeves RH, Patch D, Sharpe AH, Borriello F, Freeman GJ, Edelhoff S, Disteche C. 1997. The costimulatory genes CD80 and CD86 are linked on mouse chromosome 16 and human chromosome 3. Mamm. Genome 8:581–82 42. Ling V, Wu PW, Miyashiro JS, Marusic S, Finnerty HF, Collins M. 2001. Differential expression of inducible costimulator-ligand splice variants: lymphoid regulation of mouse GL50-b and human GL50 molecules. J. Immunol. 166:7300– 8 43. Borriello F, Freeman GJ, Edelhoff S, Disteche CM, Nadler LM, Sharpe AH. 1994. Characterization of the murine B7-1 genomic locus reveals an additional exon encoding an alternative cytoplasmic domain and a chromosomal location of chromosome 16, band B5. J. Immunol. 153:5038– 48 44. Borriello F, Oliveros J, Freeman GJ, Nadler LM, Sharpe AH. 1995. Differential expression of alternate mB7-2 transcripts. J. Immunol. 155:5490–97 45. Jeannin P, Magistrelli G, Aubry JP, Caron G, Gauchat JF, Renno T, Herbault N,

46.

47.

48.

49.

50.

51.

52.

53.

Goetsch L, Blaecke A, Dietrich PY, Bonnefoy JY, Delneste Y. 2000. Soluble CD86 is a costimulatory molecule for human T lymphocytes. Immunity 13:303–12 Aicher A, Hayden-Ledbetter M, Brady WA, Pezzutto A, Richter G, Magaletti D, Buckwalter S, Ledbetter JA, Clark EA. 2000. Characterization of human inducible costimulator ligand expression and function. J. Immunol. 164:4689–96 Ozkaynak E, Gao W, Shemmeri N, Wang C, Gutierrez-Ramos JC, Amaral J, Qin S, Rottman JB, Coyle AJ, Hancock WW. 2001. Importance of ICOS-B7RP-1 costimulation in acute and chronic allograft rejection. Nat. Immunol. 2:591–96 Yoshinaga SK, Zhang M, Pistillo J, Horan T, Khare SD, Miner K, Sonnenberg M, Boone T, Brankow D, Dai T, Delaney J, Han H, Hui A, Kohno T, Manoukian R, Whoriskey JS, Coccia MA. 2000. Characterization of a new human B7-related protein: B7RP-1 is the ligand to the costimulatory protein ICOS. Int. Immunol. 12:1439–47 Gonzalo JA, Tian J, Delaney T, Corcoran J, Rottman JB, Lora J, Al-Garawi A, Kroczek R, Gutierrez-Ramos JC, Coyle AJ. 2001. ICOS is critical for T helper cell-mediated lung mucosal inflammatory responses. Nat. Immunol. 2:597–604 Riley JL, Blair PJ, Musser JT, Abe R, Tezuka K, Tsuji T, June CH. 2001. ICOS costimulation requires IL-2 and can be prevented by CTLA-4 engagement. J. Immunol. 166:4943–48 Rottman JB, Smith T, Tonra JR, Ganley K, Bloom T, Silva R, Pierce B, Gutierrez-Ramos JC, Ozkaynak E, Coyle AJ. 2001. The costimulatory molecule ICOS plays an important role in the immunopathogenesis of EAE. Nat. Immunol. 2: 605–11 Wallin JJ, Liang L, Bakardjiev A, Sha WC. 2001. Enhancement of CD8(+) T cell responses by ICOS/B7h costimulation. J. Immunol. 167:132–39 Dong C, Temann UA, Flavell RA. 2001.

4 Feb 2002

11:14

AR

AR152-02.tex

AR152-02.SGM

LaTeX2e(2001/05/10)

P1: GSR

NEW B7-LIKE LIGANDS AND RECEPTORS

54.

Annu. Rev. Immunol. 2002.20:29-53. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

55.

56.

57.

58.

59.

60.

61.

Cutting edge: critical role of inducible costimulator in germinal center reactions. J. Immunol. 166:3659–62 Renshaw BR, Fanslow WC III, Armitage RJ, Campbell KA, Liggitt D, Wright B, Davison BL, Maliszewski CR. 1994. Humoral immune responses in CD40 liganddeficient mice. J. Exp. Med. 180:1889– 900 Xu J, Foy TM, Laman JD, Elliott EA, Dunn JJ, Waldschmidt TJ, Elsemore J, Noelle RJ, Flavell RA. 1994. Mice deficient for the CD40 ligand. Immunity 1: 423–31 Bettelli E, Das MP, Howard ED, Weiner HL, Sobel RA, Kuchroo VK. 1998. IL-10 is critical in the regulation of autoimmune encephalomyelitis as demonstrated by studies of IL-10- and IL-4deficient and transgenic mice. J. Immunol. 161:3299–306 Shevach EM. 2000. Regulatory T cells in autoimmmunity. Annu. Rev. Immunol. 18:423–49 Salomon B, Lenschow DJ, Rhee L, Ashourian N, Singh B, Sharpe A, Bluestone JA. 2000. B7/CD28 costimulation is essential for the homeostasis of the CD4+CD25+ immunoregulatory T cells that control autoimmune diabetes. Immunity 12:431–40 Chang TT, Jabs C, Sobel RA, Kuchroo VK, Sharpe AH. 1999. Studies in B7deficient mice reveal a critical role for B7 costimulation in both induction and effector phases of experimental autoimmune encephalomyelitis. J. Exp. Med. 190:733– 40 Larsen CP, Elwood ET, Alexander DZ, Ritchie SC, Hendrix R, Tucker-Burden C, Cho HR, Aruffo A, Hollenbaugh D, Linsley PS, Winn KJ, Pearson TC. 1996. Long-term acceptance of skin and cardiac allografts after blocking CD40 and CD28 pathways. Nature 381:434–38 Ensminger SM, Spriewald BM, Sorensen HV, Witzke O, Flashman EG, Bushell A, Morris PJ, Rose ML, Rahemtulla A,

62.

63.

64. 65.

66.

67.

68.

69.

70.

51

Wood KJ. 2001. Critical role for IL-4 in the development of transplant arteriosclerosis in the absence of CD40-CD154 costimulation. J. Immunol. 167:532–41 Ishida Y, Agata Y, Shibahara K, Honjo T. 1992. Induced expression of PD-1, a novel member of the immunoglobulin gene superfamily, upon programmed cell death. Embo J. 11:3887–95 Agata Y, Kawasaki A, Nishimura H, Ishida Y, Tsubata T, Yagita H, Honjo T. 1996. Expression of the PD-1 antigen on the surface of stimulated mouse T and B lymphocytes. Int. Immunol. 8:765–72 Ravetch JV, Lanier LL. 2000. Immune inhibitory receptors. Science 290:84–89 Shinohara T, Taniwaki M, Ishida Y, Kawaichi M, Honjo T. 1994. Structure and chromosomal localization of the human PD-1 gene (PDCD1). Genomics 23: 704–6 Finger LR, Pu, Wasserman R, Vibhakar R, Louie E, Hardy RR, Burrows PD, Billips LG. 1997. The human PD-1 gene: complete cDNA, genomic organization, and developmentally regulated expression in B cell progenitors. Gene 197:177– 87 Nishimura H, Agata Y, Kawasaki A, Sato M, Imamura S, Minato N, Yagita H, Nakano T, Honjo T. 1996. Developmentally regulated expression of the PD-1 protein on the surface of double-negative (CD4-CD8-) thymocytes. Int. Immunol. 8: 773–80 Nishimura H, Honjo T, Minato N. 2000. Facilitation of beta selection and modification of positive selection in the thymus of PD-1-deficient mice. J. Exp. Med. 191:891–98 Nishimura H, Nose M, Hiai H, Minato N, Honjo T. 1999. Development of lupuslike autoimmune diseases by disruption of the PD-1 gene encoding an ITIM motif-carrying immunoreceptor. Immunity 11:141–51 Nishimura H, Honjo T. 2001. PD-1: an inhibitory immunoreceptor involved in

4 Feb 2002

11:14

52

Annu. Rev. Immunol. 2002.20:29-53. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

71.

72.

73.

74.

75.

76.

77.

AR

AR152-02.tex

CARRENO

¥

AR152-02.SGM

LaTeX2e(2001/05/10)

P1: GSR

COLLINS

peripheral tolerance. Trends Immunol. 22: 265–68 Freeman GJ, Long AJ, Iwai Y, Bourque K, Chernova T, Nishimura H, Fitz LJ, Malenkovich N, Okazaki T, Byrne MC, Horton HF, Fouser L, Carter L, Ling V, Bowman MR, Carreno BM, Collins M, Wood CR, Honjo T. 2000. Engagement of the PD-1 immunoinhibitory receptor by a novel B7 family member leads to negative regulation of lymphocyte activation. J. Exp. Med. 192:1027–34 Latchman Y, Wood CR, Chernove T, Chaudhary D, Borde M, Chernova I, Iwai Y, Long AJ, Brown JA, Nunes R, Greenfield EA, Bourque K, Boussiotis VA, Carter LL, Carreno BM, Malenkovich N, Nishimura H, Honjo T, Sharpe AH, Freeman GJ. 2001. PD-L2, a novel B7 homologue, is a second ligand for PD-1 and inhibits T cell activation. Nat. Immunol. 2:261–68 Dong H, Zhu G, Tamada K, Chen L. 1999. B7-H1, a third member of the B7 family, co-stimulates T-cell proliferation and interleukin-10 secretion. Nat. Med. 5:1365–69 Tseng SY, Otsuji M, Gorski K, Huang X, Slansky JE, Pai SI, Shalabi A, Shin T, Pardoll DM, Tsuchiya H. 2001. B7-DC, a new dendritic cell molecule with potent costimulatory properties for T cells. J. Exp. Med. 193:839–46 Vermeire K, Heremans H, Vandeputte M, Huang S, Billiau A, Matthys P. 1997. Accelerated collagen-induced arthritis in IFN-gamma receptor-deficient mice. J. Immunol. 158:5507–13 Billiau A, Heremans H, Vandekerckhove F, Dijkmans R, Sobis H, Meulepas E, Carton H. 1988. Enhancement of experimental allergic encephalomyelitis in mice by antibodies against IFN-gamma. J. Immunol. 140:1506–10 Krakowski M, Owens T. 1996. Interferongamma confers resistance to experimental allergic encephalomyelitis. Eur. J. Immunol. 26:1641–46

78. Marengere LE, Waterhouse P, Duncan GS, Mittrucker HW, Feng GS, Mak TW. 1996. Regulation of T cell receptor signaling by tyrosine phosphatase SYP association with CTLA-4. Science 272:1170–73 79. Krummel MF, Allison JP. 1996. CTLA4 engagement inhibits IL-2 accumulation and cell cycle progression upon activation of resting T cells. J. Exp. Med. 183:2533– 40 80. Walunas TL, Bakker CY, Bluestone JA. 1996. CTLA-4 ligation blocks CD28dependent T cell activation. J. Exp. Med. 183:2541–50 81. Brunner MC, Chambers CA, Chan FK, Hanke J, Winoto A, Allison JP. 1999. CTLA-4-mediated inhibition of early events of T cell proliferation. J. Immunol. 162:5813–20 82. Nishimura H, Minato N, Nakano T, Honjo T. 1998. Immunological studies on PD-1 deficient mice: implication of PD1 as a negative regulator for B cell responses. Int. Immunol. 10:1563–72 83. Nishimura H, Okazaki T, Tanaka Y, Nakatani K, Hara M, Matsumori A, Sasayama S, Mizoguchi A, Hial H, Minato N, Honjo T. 2001. Autoimmune dilated cardiomyopathy in PD-1 receptor deficient mice. Science 291:319–22 84. Chambers CA, Cado D, Truong T, Allison JP. 1997. Thymocyte development is normal in CTLA-4-deficient mice. Proc. Natl. Acad. Sci. USA. 94:9296–301 85. Chapoval AI, Ni J, Lau JS, Wilcox RA, Flies DB, Liu D, Dong H, Sica GL, Zhu G, Tamada K, Chen L. 2001. B7-H3: a costimulatory molecule for T cell activation and IFN-gamma production. Nat. Immunol. 2:269–74 86. Mandelbrot DA, Oosterwegel MA, Shimizu K, Yamada A, Freeman GJ, Mitchell RN, Sayegh MH, Sharpe AH. 2001. B7-dependent T-cell costimulation in mice lacking CD28 and CTLA4. J. Clin. Invest. 107:881–87 87. Yamada A, Kishimoto K, Dong VM, Sho M, Salama AD, Anosova NG, Benichou

4 Feb 2002

11:14

AR

AR152-02.tex

AR152-02.SGM

LaTeX2e(2001/05/10)

P1: GSR

NEW B7-LIKE LIGANDS AND RECEPTORS

88.

Annu. Rev. Immunol. 2002.20:29-53. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

89.

90.

91.

92.

93.

G, Mandelbrot DA, Sharpe AH, Turka LA, Auchincloss H Jr, Sayegh MH. 2001. CD28-independent costimulation of T cells in alloimmune responses. J. Immunol. 167:140–46 Linsley PS, Peach R, Gladstone R, Bajorath J. 1994. Extending the B7 (CD80) gene family. Protein Sci. 3:1341–43 Henry J, Miller MM, Pontarotti P. 1999. Structure and evolution of the extended B7 family. Immunol. Today 20:285– 88 Mather IH, Jack LJ. 1993. A review of the molecular and cellular biology of butyrophilin, the major protein of bovine milk fat globule membrane. J. Dairy Sci. 76:3832–50 Taylor MR, Peterson JA, Ceriani RL, Couto JR. 1996. Cloning and sequence analysis of human butyrophilin reveals a potential receptor function. Biochim. Biophys. Acta. 1306:1–4 Ruddy DA, Kronmal GS, Lee VK, Mintier GA, Quintana L, Domingo R Jr, Meyer NC, Irrinki A, McClelland EE, Fullan A, Mapa FA, Moore T, Thomas W, Loeb DB, Harmon C, Tsuchihashi Z, Wolff RK, Schatzman RC, Feder JN. 1997. A 1.1-Mb transcript map of the hereditary hemochromatosis locus. Genome Res. 7:441–56 Rhodes DA, Stammers M, Malcherek G,

94.

95.

96.

97.

53

Beck S, Trowsdale J. 2001. The cluster of BTN genes in the extended major histocompatibility complex. Genomics 71: 351–62 Stefferl A, Schubart A, Storch M, Amini A, Mather I, Lassmann H, Linington C. 2000. Butyrophilin, a milk protein, modulates the encephalitogenic T cell response to myelin oligodendrocyte glycoprotein in experimental autoimmune encephalomyelitis. J. Immunol. 165: 2859–65 Pham-Dinh D, Mattei MG, Nussbaum JL, Roussel G, Pontarotti P, Roeckel N, Mather IH, Artzt K, Lindahl KF, Dautigny A. 1993. Myelin/oligodendrocyte glycoprotein is a member of a subset of the immunoglobulin superfamily encoded within the major histocompatibility complex. Proc. Natl. Acad. Sci. USA 90:7990– 94 Pham-Dinh D, Jones EP, Pitiot G, Della Gaspera B, Daubas P, Mallet J, Le Paslier D, Fischer Lindahl K, Dautigny A. 1995. Physical mapping of the human and mouse MOG gene at the distal end of the MHC class Ib region. Immunogenetics 42:386–91 Fahrer AM, Bazan JF, Papathanasiou P, Nelms KA, Goodnow CC. 2001. A genomic view of immunology. Nature 409: 836–38

P1: FRK

February 1, 2002

12:47

Annual Reviews

AR152-FM

Annual Review of Immunology Volume 20, 2002

Annu. Rev. Immunol. 2002.20:29-53. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

CONTENTS FRONTISPIECE, Charles A. Janeway, Jr. A TRIP THROUGH MY LIFE WITH AN IMMUNOLOGICAL THEME, Charles A. Janeway, Jr.

xiv 1

THE B7 FAMILY OF LIGANDS AND ITS RECEPTORS: NEW PATHWAYS FOR COSTIMULATION AND INHIBITION OF IMMUNE RESPONSES, Beatriz M. Carreno and Mary Collins

29

MAP KINASES IN THE IMMUNE RESPONSE, Chen Dong, Roger J. Davis, and Richard A. Flavell

PROSPECTS FOR VACCINE PROTECTION AGAINST HIV-1 INFECTION AND AIDS, Norman L. Letvin, Dan H. Barouch, and David C. Montefiori T CELL RESPONSE IN EXPERIMENTAL AUTOIMMUNE ENCEPHALOMYELITIS (EAE): ROLE OF SELF AND CROSS-REACTIVE ANTIGENS IN SHAPING, TUNING, AND REGULATING THE AUTOPATHOGENIC T CELL REPERTOIRE, Vijay K. Kuchroo, Ana C. Anderson, Hanspeter Waldner, Markus Munder, Estelle Bettelli, and Lindsay B. Nicholson

55 73

101

NEUROENDOCRINE REGULATION OF IMMUNITY, Jeanette I. Webster, Leonardo Tonelli, and Esther M. Sternberg

125

MOLECULAR MECHANISM OF CLASS SWITCH RECOMBINATION: LINKAGE WITH SOMATIC HYPERMUTATION, Tasuku Honjo, Kazuo Kinoshita, and Masamichi Muramatsu

165

INNATE IMMUNE RECOGNITION, Charles A. Janeway, Jr. and Ruslan Medzhitov

KIR: DIVERSE, RAPIDLY EVOLVING RECEPTORS OF INNATE AND ADAPTIVE IMMUNITY, Carlos Vilches and Peter Parham ORIGINS AND FUNCTIONS OF B-1 CELLS WITH NOTES ON THE ROLE OF CD5, Robert Berland and Henry H. Wortis E PROTEIN FUNCTION IN LYMPHOCYTE DEVELOPMENT, Melanie W. Quong, William J. Romanow, and Cornelis Murre

197 217 253 301

LYMPHOCYTE-MEDIATED CYTOTOXICITY, John H. Russell and Timothy J. Ley

SIGNAL TRANSDUCTION MEDIATED BY THE T CELL ANTIGEN x

323

P1: FRK

February 1, 2002

12:47

Annual Reviews

AR152-FM

CONTENTS

xi

RECEPTOR: THE ROLE OF ADAPTER PROTEINS, Lawrence E. Samelson INTERACTION OF HEAT SHOCK PROTEINS WITH PEPTIDES AND ANTIGEN PRESENTING CELLS: CHAPERONING OF THE INNATE AND ADAPTIVE IMMUNE RESPONSES, Pramod Srivastava CHROMATIN STRUCTURE AND GENE REGULATION IN THE IMMUNE SYSTEM, Stephen T. Smale and Amanda G. Fisher PRODUCING NATURE’S GENE-CHIPS: THE GENERATION OF PEPTIDES FOR DISPLAY BY MHC CLASS I MOLECULES, Nilabh Shastri, Susan

371

Annu. Rev. Immunol. 2002.20:29-53. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

Schwab, and Thomas Serwold

395 427

463

THE IMMUNOLOGY OF MUCOSAL MODELS OF INFLAMMATION, Warren Strober, Ivan J. Fuss, and Richard S. Blumberg T CELL MEMORY, Jonathan Sprent and Charles D. Surh

GENETIC DISSECTION OF IMMUNITY TO MYCOBACTERIA: THE HUMAN MODEL, Jean-Laurent Casanova and Laurent Abel ANTIGEN PRESENTATION AND T CELL STIMULATION BY DENDRITIC CELLS, Pierre Guermonprez, Jenny Valladeau, Laurence Zitvogel, Clotilde Th´ery, and Sebastian Amigorena

495 551 581

621

NEGATIVE REGULATION OF IMMUNORECEPTOR SIGNALING, Andr´e Veillette, Sylvain Latour, and Dominique Davidson

669

CPG MOTIFS IN BACTERIAL DNA AND THEIR IMMUNE EFFECTS, Arthur M. Krieg

PROTEIN KINASE Cθ

709 IN

T CELL ACTIVATION, Noah Isakov and Amnon

Altman

RANK-L AND RANK: T CELLS, BONE LOSS, AND MAMMALIAN EVOLUTION, Lars E. Theill, William J. Boyle, and Josef M. Penninger PHAGOCYTOSIS OF MICROBES: COMPLEXITY IN ACTION, David M. Underhill and Adrian Ozinsky

761 795 825

STRUCTURE AND FUNCTION OF NATURAL KILLER CELL RECEPTORS: MULTIPLE MOLECULAR SOLUTIONS TO SELF, NONSELF DISCRIMINATION, Kannan Natarajan, Nazzareno Dimasi, Jian Wang, Roy A. Mariuzza, and David H. Margulies

853

INDEXES Subject Index Cumulative Index of Contributing Authors, Volumes 1–20 Cumulative Index of Chapter Titles, Volumes 1–20

ERRATA An online log of corrections to Annual Review of Immunology chapters (if any, 1997 to the present) may be found at http://immunol.annualreviews.org/

887 915 925

8 Feb 2002

9:33

AR

AR152-03.tex

AR152-03.SGM

LaTeX2e(2001/05/10) P1: GSR 10.1146/annurev.immunol.20.091301.131133

Annu. Rev. Immunol. 2002. 20:55–72 DOI: 10.1146/annurev.immunol.20.091301.131133 c 2002 by Annual Reviews. All rights reserved Copyright °

MAP KINASES IN THE IMMUNE RESPONSE Chen Dong,1 Roger J. Davis,2 and Richard A. Flavell3

Annu. Rev. Immunol. 2002.20:55-72. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

1

Department of Immunology, University of Washington School of Medicine, Seattle, Washington 98195-7650; e-mail: [email protected] 2 Howard Hughes Medical Institute and Program in Molecular Medicine, Department of Biochemistry and Molecular Biology, University of Massachusetts Medical School, Worcester, Massachuetts 01605; e-mail: [email protected] 3 Howard Hughes Medical Institute and Section of Immunobiology, Yale University School of Medicine, New Haven, Connecticut 06520; e-mail: [email protected]

Key Words JNK, p38, ERK, innate immunity, helper T Cells ■ Abstract MAP kinases are among the most ancient signal transduction pathways and are widely used throughout evolution in many physiological processes. In mammalian species, MAP kinases are involved in all aspects of immune responses, from the initiation phase of innate immunity, to activation of adaptive immunity, and to cell death when immune function is complete. In this review, we summarize recent progress in understanding the function and regulation of MAP kinase pathways in these phases of immune responses.

INTRODUCTION Immune responses involve a number of cell types that function as initiators, regulators, and effectors. These cells interact with and cross-regulate each other, and the target cells respond using signal transduction pathways to mediate gene expression and immune function. The MAP kinase cascade is one of the most ancient and evolutionarily conserved signaling pathways, which is also important for many processes in immune responses. There are three major groups of MAP kinases in mammalian cells—the extracellular signal-regulated protein kinases (ERK) (1), the p38 MAP kinases (2), and the c-Jun NH2-terminal kinases (JNK) (3, 4) (Figure 1). These MAP kinases are activated by dual phosphorylation at the tripeptide motif Thr-Xaa-Tyr. The sequence of this tripeptide motif is different in each group of MAP kinases: ERK (Thr-Glu-Tyr); p38 (Thr-Gly-Tyr); and JNK (Thr-Pro-Tyr). The dual phosphorylation of Thr and Tyr is mediated by a conserved protein kinase cascade. The ERK MAP kinases are activated by the MAP kinase kinases (MKK) MKK1 and MKK2; the p38 MAP kinases are activated by MKK3, MKK4, and MKK6; and the JNK pathway is activated by MKK4 and MKK7. These MAP kinase kinases are activated, in turn, by several different MAP kinase kinase kinases (MKKK). 0732-0582/02/0407-0055$14.00

55

1 Feb 2002

10:8

Annu. Rev. Immunol. 2002.20:55-72. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

56

AR

AR152-03.tex

DONG

¥

DAVIS

¥

AR152-03.SGM

LaTeX2e(2001/05/10)

P1: GSR

FLAVELL

Figure 1 Mammalian MAP kinase pathways.

Different upstream signals can lead to the activation of these MKKK. However, prominent roles for small G proteins have been identified. Thus, the ERK pathway can be activated by Ras via the Raf group of MKKK. In contrast, the p38 and JNK MAP kinases are activated by Rho family GTPases, including Rac and Cdc42. Candidate MKKK that are activated by Rho proteins include members of the MEKK and mixed-lineage protein kinase (MLK) groups. Signal transduction along the MAP kinase pathways can be facilitated by scaffold proteins. The JIP group of proteins has been recently identified as scaffolds for the JNK pathway, which connects MLK to MKK7 and to JNK [recently reviewed by Davis (3)]. MAP kinases play important functions in lymphocyte development and have been well reviewed recently (5–7). In this review, we focus on recent progress on the function and regulation of MAP kinases in different components or phases of immune responses, with particular focus on the studies achieved through mouse genetic approaches. Due to space limitations, we apologize that we are not able to cover all aspects of MAP kinase research here.

MAP KINASES IN INNATE IMMUNE RESPONSES Immune responses to foreign organisms can be traced back to invertebrates. In the fruit fly, Drosophila, a group of 8 proteins known as the Toll family is central to innate defense, which involves the recognition of bacteria and fungi, and the induction of antimicrobial peptides to kill an invader (8). It is well established that the

1 Feb 2002

10:8

AR

AR152-03.tex

AR152-03.SGM

LaTeX2e(2001/05/10)

P1: GSR

Annu. Rev. Immunol. 2002.20:55-72. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

MAP KINASES

57

NF-κB pathway is an important downstream target of Toll receptors in Drosophila innate immune responses. MAP kinases such as JNK and p38 are activated by pathogens in insect cells (9, 10). Fly p38 appears to attenuate antimicrobial peptide gene expression following exposure to lipopolysaccharide (LPS) (9). The role of JNK in Drosophila innate immune response has not been defined genetically due to the embryonic lethality of the flies deficient in components of this pathway. In mammalian systems, 10 Toll-like receptors (TLR) exist. Recent years have seen rapid progress in defining the role of these TLRs in recognition of “pattern” molecules on microbial organisms and, as a consequence, induction of inflammatory responses (11). For instance, TLR4 is critical for LPS recognition (12–14); TLR2 for lipoteichoic acids (LTAs) and peptidoglycan (PGN) (15, 16); TLR5 for bacterial flagellin (17); and TLR9 for CpG-containing DNA (18). TLR cytoplasmic domains resemble that of the interleukin 1 (IL-1) receptor, and hence they are called Toll-IL-1-Receptor (TIR) domains. After ligand-mediated dimerization, TLR recruits an adaptor protein MyD88 (19–24). MyD88 then assembles a signalsome containing IRAK, TRAF6, and ECSIT (24–26). Together, they mediate activation of NF-κB and MAP kinases such as p38 and JNK, which leads to the production of inflammatory cytokines such as tumor necrosis factor (TNF) α, IL-1, and IL-12. In support of the roles of these signaling intermediates in mediating MAP kinase activation, genetic ablation of MyD88 resulted in defective MAP kinase activation in response to endotoxin (19), and IRAK−/− and TRAF6−/− cells exhibited deficiency of p38 and JNK activation in response to IL-1 (27, 28).

MKK3-p38 Mediates IL-12 Production in Innate Immune Responses To examine the functional roles of p38 MAP kinase in immune responses, Lu et al. analyzed mice deficient in the p38 kinase MKK3 (29). Mkk3 disruption caused an approximately 40%–70% reduction of total p38 activity in macrophages treated with LPS. The residual activity could be contributed by MKK6 or perhaps MKK4. Mkk3−/− macrophages exhibited a selective defect in LPS-induced IL-12 production at both protein and RNA levels. The same results were achieved by using p38-specific inhibitors SB 203580, SB 202190, and SB 202474. Production of other cytokines such as TNFα, IL-6, IL-1α, and IL-1α were comparable between wild-type and knockout cells, and addition of p38 inhibitors appeared to have an additive inhibitory effect only on IL-1α and IL-1β expression by knockout cells, but not on TNF or IL-6 expression. The MKK3-p38 pathway seems therefore to play a specific role in the activation of IL-12 production in macrophages; p38 activation by other kinases may compensate during the IL-1 regulation. The essential function of MKK3 in IL-12 regulation was reported in the same study to exist in dendritic cells (DC). CD40-CD40L interaction, a receptor-ligand pair upstream of p38 and JNK, is the main activator of IL-12 production in these cells. MKK3 deficiency resulted in a severe reduction of IL-12 secretion by bone marrow–derived DC activated by CD40L.

1 Feb 2002

10:8

58

AR

AR152-03.tex

DONG

¥

DAVIS

¥

AR152-03.SGM

LaTeX2e(2001/05/10)

P1: GSR

FLAVELL

Annu. Rev. Immunol. 2002.20:55-72. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

Regulation of TNFα Production by MAP Kinases p38 MAP kinase activates many protein kinases such as MAP-kinase-activated protein kinases (MAPKAP) MK2 and MK3. Kotlyarov et al. generated and analyzed mice deficient for MAPKAP MK2 (30). These mice showed increased stress resistance and survived LPS shock. Interestingly, production of TNFα protein in the serum of LPS-treated animals in vivo, and by LPS-treated splenocytes in vitro, was greatly compromised, whereas the mRNA levels were not affected. Therefore, these investigators proposed that MAPKAP MK2 regulates TNFα biosynthesis at the posttranscriptional level. p38 and MAPKAP MK2 may exert their regulation of TNFα translation through AU-rich elements (ARE) in the 30 untranslated region of TNFα transcripts. Targeted deletion of these elements in the mouse genome resulted in abnormal TNFα gene expression and development of two types of autoimmunity, i.e., inflammatory arthritis and inflammatory bowel diseases (31). While p38 inhibitors blocked TNFα production in normal mice, ARE-deleted mice did not respond to the drug, indicating a requirement for the ARE region in the p38-mediated activation of TNF translation. In addition, IL-10, an anti-inflammatory cytokine, was recently found to repress TNFα production by the inhibition of p38-MAPKAP MK2-mediated ARE activity (32). Evidence for ERK regulation of TNFα induction was also recently reported. Tpl2/Cot is a proto-oncogene that serves as a MAP kinase kinase kinase (MKKK) (33, 34). Mice deficient for this kinase were found deficient in TNFα production when exposed to LPS (35). The macrophages from these mice exhibited selective ERK activation deficits, and the ERK inhibitor PD98059 had a similar effect. Deletion of the ARE motif in TNFα mRNA minimized the effect of Tpl2 deficiency, suggesting that ERK may target the ARE region as well. However, ERK may work at a different phase than p38/MAPKAP MK2. TNFα mRNA transport from the nucleus to the cytoplasm was inhibited by ERK inhibitor or Tpl2 inactivation.

Cellular Responses to Inflammatory Cytokines: The Roles of p38 and JNK p38 and JNK MAP kinases are preferentially activated by inflammatory cytokines such as TNF and IL-1, and they have critical functions in cellular responses to these cytokines. p38 MKK3-mediated p38 activity was reported by Wysk et al. to be important also for TNF-induced cytokine production (36). p38 activation was selectively inhibited in Mkk3−/− mouse embryonic fibroblasts (MEF) in response to TNFα, but not to IL-1, UV, or sorbitol. As a result of this deficiency, Mkk3−/− MEF failed to upregulate IL-1α, IL-1β, IL-6, and TNFα mRNA or to downregulate IL-1 receptor antagonist (RA) in response to TNF. This study clearly indicates the essential function of MKK3-p38 in the cellular response to TNF, but not to

1 Feb 2002

10:8

AR

AR152-03.tex

AR152-03.SGM

LaTeX2e(2001/05/10)

P1: GSR

MAP KINASES

59

Annu. Rev. Immunol. 2002.20:55-72. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

IL-1. However, p38, especially p38α, does play an important role in IL-1-mediated inflammatory response. Allen et al. examined p38α −/− MEF and found that IL-1induced IL-6 production was greatly compromised (37). It is still unclear which MKK may mediate p38 activation in IL-1 signal transduction. Given the absence of deficiency in Mkk3−/− MEF, then MKK6, or MKK4 are candidates. It is also possible that these three MKKs function redundantly. To support a role for MKK4 in p38 activation, Ganiatsas et al. reported that Mkk4−/− MEF exhibited profound reduction of p38 activation in response to IL-1 but not to anisomycin, although JNK activation is affected in both cases (38). JNK activity can be strongly induced in multiple cell types by liposaccharides (LPS) or inflammatory cytokines such as TNF and IL-1 (3, 4). It is interesting that Drosophila cells activate the JNK pathway in response to LPS (10), suggesting that the JNK pathway is evolutionarily conserved in innate immune responses. However, the functional roles of JNK have not been elucidated in these systems. In fibroblasts, JNK can also be activated by inflammatory cytokines and doublestranded viral RNA (39). Using JNK2−/− fibroblasts, Chu et al. demonstrated that JNK is required for production of multiple cytokines including type I interferon and IL-6 (39). It is intriguing that JNK1 did not compensate in the above experiments. Recently, Han et al. studied the role of JNK in rheumatoid arthritis fibroblast-like synoviocytes (40). By use of a novel JNK-specific inhibitor and synoviocytes from JNK1 or JNK2 knockout mice, they reported that JNK is required for IL-1-mediated collagenase-3 expression and joint inflammation. JNK is activated by TRAF2 in the TNF signaling pathway and by TRAF6 following IL-1 activation (27, 41). Further downstream, one JNK kinase—MKK7, but not the other—MKK4, appears to play an essential function in activating JNK in TNF- or IL-1-treated MEF, although both are involved in JNK activation in response to physical stress (42). Tournier et al. examined JNK activation in MEF deficient in either MKK4 or MKK7, or both; they found that MKK7 is solely required for JNK activation in these cells in response to inflammatory cytokines. It is interesting that MKK4 and MKK7 preferentially phosphorylate different target amino acids of the tripeptide motif in JNK, whereas MKK4 phosphorylates tyrosine, MKK7 phosphorylates threonine. JNK activation by MKK7 appears to be an essential part of inflammatory responses. JNK

MAP KINASES IN HELPER T CELL ACTIVATION AND DIFFERENTIATION CD4+ helper T (Th) cells play a central regulatory role in immune responses. Like CD8 cytotoxic T cells, Th cells develop in the thymus, with specificity to recognize specific MHC-peptide complexes on antigen-presenting cells (APC). Following the receipt of signals from these innate immune cells through the T cell receptor and CD28 costimulator, Th cells are triggered to produce interleukin 2 (IL-2) and

1 Feb 2002

10:8

Annu. Rev. Immunol. 2002.20:55-72. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

60

AR

AR152-03.tex

DONG

¥

DAVIS

¥

AR152-03.SGM

LaTeX2e(2001/05/10)

P1: GSR

FLAVELL

Figure 2 The role of MAP kinase pathways in Th cell differentiation and cytokine production.

enter the cell cycle. Following or during several days of rapid cell division, these T cells differentiate into effector cells (Figure 2). During this process of differentiation, a new pattern of cytokine expression is established that provides the gene products responsible for the specific effector functions of these cells, and therefore for their ability to protect the host from a variety of pathogens. Two classes of effector CD4 T cells have been defined on the basis of the cytokines that they secrete and the immunomodulatory effects conferred by these cytokines (43–45). Effector Th1 cells produce proinflammatory cytokines such as interferon-γ (IFN-γ ) and lymphotoxin-α (Lt-α). These cytokines organize inflammatory centers and enhance cellular immune response; moreover, intracellular pathogens such as Mycobacterium, Salmonella, and other intravesicular agents are killed by IFN-γ through the activation of antimicrobial defenses. Th1 cytokine production is also characteristic of many organ-specific autoimmune diseases, including rheumatoid arthritis, insulin-dependent diabetes mellitus, experimental autoimmune encephalomyelitis (EAE), etc. Effector Th2 cells, in contrast, produce different cytokines (IL-4, IL-5, IL-9, IL-10, IL-13, and so on) that together instruct B cells to proliferate and differentiate into antibody-secreting plasma cells, as well as potentiate the function of several cell types in the antiparasite responses. As such, Th2 cells play an important role in the provision of protection against certain extracellular pathogens such as bacteria and a variety of parasites, and they are also involved in asthmatic reactions. Proper differentiation of na¨ıve Th cells into Th1 or Th2 cells is critical for a T-dependent immune response.

1 Feb 2002

10:8

AR

AR152-03.tex

AR152-03.SGM

LaTeX2e(2001/05/10)

P1: GSR

MAP KINASES

61

Leishmania major infection in mice represents one of best-studied animal models for Th differentiation and function. Most common inbred strains can mount a Th1 response and resolve the lesions; Balb/c mice, however, develop an infection that never heals and a Th2 response that renders them susceptible to the infection. On the other hand, immunodeviation of Th1 responses to the Th2 direction has been proposed as a means to alleviate the symptoms in autoimmune diseases.

Annu. Rev. Immunol. 2002.20:55-72. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

ERK Regulation of T Cell Activation and Differentiation The ERK pathway was first identified downstream of oncogenic Ras and is often involved in the regulation of cell growth and differentiation. There are two isoforms of ERK, ERK1 and 2, which are sometimes referred as p44/p42 MAP kinases. They can be activated by MEK1 and MEK2 upstream kinases. ERK activation is an important event of T cell activation. TcR engagement leads to the recruitment of multimolecular components to the cell surface, including adaptor proteins SLP76 and Grab2, which in turn activate the sos-ras-MEK-ERK pathway (46). Deficient ras and ERK activation was reported to exist in clones that are anergized, i.e., stimulated without CD28 costimulation (47, 48). However, there is also evidence that ERK inhibition alone had no influence on anergy induction (49); thus, it is unclear whether ERK deficiency is merely the result of anergy, or together with other pathways it contributes to anergization. Although there is a strong consensus on the role of ERK pathway in thymocyte selection—ERK1-deficient mice exhibited defective thymocyte maturation (50)—its function in peripheral Th cell differentiation was not studied until recently. Using dominant H-RAS transgenic mice in which ERK activation by TcR was severely compromised, Yamashita and colleagues showed that this pathway is required for Th2 differentiation (51). Similar results were found using wild-type cells treated with inhibitors against MEKs. They went further to show that the ERK pathway functions to enhance IL-4-induced STAT6 and IL-4R phosphorylation, which suggests a mechanism of cross-regulation among different signaling pathways.

p38 in Th1 Differentiation The p38 MAP kinase pathway was first reported by Rinc´on et al. to be selectively activated in mouse Th1 effector cells (52). Recent studies also suggest the role of pro-inflammatory cytokines IL-12 and IL-18 in p38 activation in T cells (53, 54). Growing evidence supports the role of the p38 pathway in Th1 differentiation and cytokine production. Imidazole inhibitors of the p38 kinases block IFN-γ production by Th1 cells in a dose-dependent manner but have no effect on IL-4 production by Th2 cells (52). Furthermore, transgenic mice in which a dominant negative p38α transgene was directed by the lck distal promoter showed reduced IFN-γ cytokine secretion and mRNA production (52). T cells from mice deficient in the p38 upstream kinase MKK3 have a defect in IFN-γ production, even when provided antigen-presenting cells from a wild-type B6 mouse (29). The effect

1 Feb 2002

10:8

Annu. Rev. Immunol. 2002.20:55-72. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

62

AR

AR152-03.tex

DONG

¥

DAVIS

¥

AR152-03.SGM

LaTeX2e(2001/05/10)

P1: GSR

FLAVELL

of the p38 pathway on IFNγ is probably transcriptional since IFN-γ promoter reporters are also inhibited by dominant negative p38-α in Jurkat cells, indicating that p38 regulation is necessary for IFNγ expression. On the other hand, transgene-encoded constitutively active MKK6, one of the upstream MAP kinase kinases that activates p38 kinase, led to the activation of p38 MAP kinase and the consequent activation of IFN-γ transcription to higher levels than seen in control transgene negative cells (52). Candidate downstream targets of p38 regulation are likely to include transcription factors of the ATF family. Studies of the IFN-γ promoter, the prototype Th1 specific cytokine, led to the identification of c-jun/ATF2 sites and a series of other ATF binding sites within two functionally active elements called the proximal and distal IFN-γ elements (55, 56). Transgenic mice in which these two elements were linked to luciferase reporter constructs (57, 58) showed that the proximal element that carries a c-jun/ATF2 site, factors which are p38 targets, exhibited Th1 specificity, whereas the distal element did not. However, ATF2-mutant cells did not exhibit an IFNγ defect (59); therefore, it is not clear at this stage whether other ATF2 isoforms or ATF factors may play compensatory roles.

JNK Regulation of Th Differentiation Three JNK-encoding genes have been identified in mammals: JNK1, 2, and 3 (3). JNK3 is selectively expressed in neuronal tissues, while JNK1 and 2 are ubiquitously expressed (3). Surprisingly, however, when examined by Northern blot analysis, expression of JNK1 and JNK2 in peripheral lymphoid tissues was only weakly detectable (60). The same results were found for the JNK kinases, MKK4 and MKK7 (42). Consistent with these results, JNK activity in na¨ıve T cells and in T cells activated for a short time (<30 min) was very low (60, 61). Both JNK expression and activity are significantly increased after T cell activation, and generally peak about 36–60 h post activation (60, 61). Activation of the TCR is required, but it is not clear what signals regulate the greatly enhanced expression and activity of JNK proteins. Interestingly, although CD28 costimulation appears to maximize JNK activity (60, 61), it is not essential for increased expression of JNK mRNA (60). In the effector phase, JNK1 and JNK2 are expressed in approximately equal amounts in Th1 and Th2 cells; however, Th1 cells have highly inducible JNK activity compared to Th2 cells (16, 62). This suggests that JNK may have important functional roles in Th1 cells. In order to understand the role of JNK in T cell activation and cytokine production, two groups generated and analyzed mice deficient in JNK2 (63, 64). B and T lymphocytes develop normally in the absence of JNK2 (63, 64). When Yang et al. examined the activation and IL-2 expression of JNK2−/− Th cells, they did not observe a defect (64) in purified CD4 T cells, although deficiencies were subsequently reported by another group using mixed cell populations (63). Sabapathy et al. reported that Jnk2−/− produced reduced IL-2 under suboptimal stimulation; indeed, JNK may regulate IL-2 mRNA transcription or stability (65, 66). But, in

1 Feb 2002

10:8

AR

AR152-03.tex

AR152-03.SGM

LaTeX2e(2001/05/10)

P1: GSR

Annu. Rev. Immunol. 2002.20:55-72. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

MAP KINASES

63

agreement with the previous group, the role of JNK2 in IL-2 expression appears nonessential in their studies because the Jnk2−/− cells secreted normal amounts of IL-2 when given higher levels of stimulation. Despite the lack of an essential role of JNK2 for IL-2 expression, JNK2 plays a critical role in Th cell differentiation and cytokine production. The inducible JNK activity in Th1 cells is inhibited in the absence of JNK2 (64), suggesting that JNK2 accounts for much of this activity. How JNK2 is activated and functions in a Th1-specific fashion is not fully established (but see below for a discussion of two candidates, Rac2 and GADD45). Yang et al. found that Jnk2−/− Th2 cells produced significantly decreased amounts of IFNγ (64), a signature cytokine for Th1 cells. Reduced levels of IFN-γ appear to derive from reduced production of the IL-12 receptor β chain (IL-12rβ2). This deficiency is a consequence of the failure to produce optimal levels of IFN-γ early during the response to T cell receptor and costimulatory receptor ligation. Addition of IFNγ to JNK2-deficient T cells restores the deficit and confers normal Th1 function on these cells in vitro. In contrast, the Th2 responses of JNK2-deficient mice appear relatively normal. Dong et al. generated and analyzed mice deficient for JNK1 (61). Interestingly, although JNK2 is present, total JNK activity was reduced in JNK1−/− Th cells activated by anti-CD3 with or without anti-CD28 for 24–60 h (61). In spite of the JNK deficiency in JNK1−/− Th cells, IL-2 was expressed normally (61). JNK1-deficient mice also exhibited deficiencies in Th differentiation; in contrast to JNK2−/− mice, they showed an exaggerated Th2 response (61). Even when cultured under Th1 conditions, Th2 cytokines IL-4, IL-5, and IL-10 are produced in significant amounts. Consistent with exaggerated Th2 responses, infection of JNK1 mice with Leishmania leads to greatly exacerbated disease with failure to heal skin lesions; the disease advances to ulceration in a manner similar to the BALB/c mouse, which also has a profoundly potent Th2 response (67). Most notably, Jnk1−/− CD4 T cells produced Th2 cytokines in the absence of CD28 costimulation and differentiate preferentially into Th2 cells in vitro when stimulated with anti-CD3, irradiated APC, and IL-2, whereas the wild-type CD4 T cells became mostly Th1 cells (61). This is likely caused by enhanced IL-4 production at the early phase (i.e., 24-h and 48-h time-points) of Th cell activation. Interestingly, in correlation with a role of JNK in inhibition of Th2 differentiation, Lieberson et al. found recently that TRAF2 also represses Th2 responses (68). TRAF2 was previously shown to be essential JNK activator in response to TNF (41); it may also mediate JNK activation in T cells. To understand how JNK negatively regulates Th2 differentiation, Dong et al. examined Th2 transcription factors expression and showed that JNK1-deficient mice have elevated NFATc in the nucleus (61). The mechanism whereby NFATc accumulates has now been elucidated (69). Specifically, Chow et al. found that JNK serves as an NFATc kinase and phosphorylates NFATc on two serine residues, one of which lies in the calcineurin binding site. JNK-mediated phosphorylation inhibits calcineurin phosphatase targeting to NFATc, thereby opposing NFATc

1 Feb 2002

10:8

Annu. Rev. Immunol. 2002.20:55-72. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

64

AR

AR152-03.tex

DONG

¥

DAVIS

¥

AR152-03.SGM

LaTeX2e(2001/05/10)

P1: GSR

FLAVELL

nuclear translocation; in the absence of JNK signaling, NFATc is inefficiently removed from the nucleus, resulting in nuclear NFATc accumulation and excessive Th2 cytokines. The role of JNK in T cell activation and IL-2 expression is relatively controversial. One group used purified CD4 T cells from JNK1- or JNK2-deficient animals to demonstrate their distinct function in Th cell differentiation and cytokine production, but they failed to support a role of JNK in T cell activation and IL-2 expression (61, 64). In contrast, another group reported an IL-2 expression defect in mixed lymphocyte populations of JNK2−/− mice (63) and, most recently, in their Jnk1−/− mice (70). These investigators described a deficiency in NFAT binding to the target sites in JNK1- or JNK2-deficient cells treated with maximal stimuli, although IL-2 expression appeared normal under such conditions (63, 70). Both JNK1 and JNK2 appeared to have an identical function in these studies with neither enzyme compensating for the deficiency of the other in their single-gene knockout models. However, JNK1 and JNK2 appear to play overlapping roles in embryonic development, since JNK1- or JNK2-deficient animals are viable, although the double knockout mice die as early embryos (71, 72). It is difficult to reconcile these very different results with apparently equivalent mutant mice. Possible explanations include the use of different cell populations [purified CD4 T cells in (61, 64) compared with a mixture of CD4, CD8 T cells, B cells, and myeloid cells in (63, 70)] as well as possible differences in genetic background. Further study is required to resolve this. To investigate the possible redundant function of JNK1 and JNK2 in T cells, which might account for the difference in the published reports, animals deficient in both JNK1 and 2 in CD4 T cells were studied (73). Dong et al. first bred transgenic mice that express dominant negative JNK1 (dnJNK1) in T cells with JNK2−/− mice, and secondly prepared JNK1−/− JNK2−/− embryonic stem cells and used them to reconstitute Rag1-deficient blastocysts. Th cells from both animals (for simplicity, we call them JNK-deficient cells) were found to have no deficiency in IL-2 production 24 h after activation with anti-CD3 in the absence or presence of anti-CD28. In contrast, JNK-deficient CD4 T cells even produced two- to threefold more IL-2 than wild-type control cells. Increased IL-2 production was also found in JNK-deficient Th cells treated with low doses of stimulation. Consistent with the increased IL-2 production, JNK-deficient Th cells exhibited greater proliferation (73). In keeping with the JNK1−/− cells, Th cells deficient for both JNK1 and JNK2 were found to produce exaggerated amounts of Th2 cytokines and preferentially developed along the Th2 lineage (73). These results resolved the role of JNK in activation of primary mouse CD4 T cells: JNK is not required for T cell activation and IL2 expression; instead it is essential for Th differentiation and effector cytokine production. This function correlates with the timing of the expression of JNK signaling components: low expression in na¨ıve and recently activated cells, but highly induced in effector Th1 cells. Thus, we suggest a new model for the role of JNK in T cell immune responses: to reduce proliferative responses of the activated Th cells and to potentiate their polarized T cell differentiation into the Th1 lineage.

1 Feb 2002

10:8

AR

AR152-03.tex

AR152-03.SGM

LaTeX2e(2001/05/10)

P1: GSR

MAP KINASES

65

Annu. Rev. Immunol. 2002.20:55-72. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

Coordinated Regulation of JNK and p38 in Th1 Cells- Rac2 and GADD45 p38 and JNK MAP kinases are frequently coordinately regulated in many physiological processes. For instance, Grb2 haplo-deficiency resulted in deficient p38 and JNK activation in thymocytes, while ERK remains normal (74). As a result, thymocyte negative selection is selectively reduced. This study suggests that both p38 and JNK require a higher threshold of activation signals. In Th1 cells, p38 and JNK MAP kinases are both selectively activated; the mechanism that underlies this selectivity was, until recently, not understood. Li et al. found that the expression of the hemopoietic cell–specific small G protein Rac2 is enhanced in Th1 cells (75). Since Rac2 is a likely upstream effector for these MAP kinase pathways, Li et al. tested the hypothesis that elevated Rac2 mediated this selective activation. Indeed, Rac2 activation drives IFNγ transcription in vitro and in vivo in Rac2 transgenic mice through several signaling pathways, including p38 and NF-κB. Blocking the Rac2 pathway led to the inhibition of IFNγ production and gene transcription (75). T cell deficiency in Rac2 also leads to reduced IFNγ production (75). Another factor involved in JNK signaling with selective Th1 expression is GADD45. Previous studies have established that GADD45γ can bind MEKK4/ MTK1 and activate stress-activated MAPK in transfection assays (76). Since MEKK4/MTK1 has a CRIB domain that binds Rac, it is possible that MEKK4/ MTK1 may integrate signals from both GADD45 proteins. Two isoforms of GADD45, GADD45β and GADD45γ, are induced in Th1 cells, by IL-12 and IL-18 (53, 77). Interestingly, IL-2 specifically upregulates GADD45γ (78). Th1 cells from GADD45γ −/− mice were unable to activate p38 and JNK in response to TCR signaling and produced much less IFN-γ upon restimulation (77). Likewise, GADD45β overexpression increases IFN-γ production (53). GADD45β mediates cytokine (IL-12 and IL-18) activated IFN-γ production via the p38 pathway (53), and this pathway is deficient in GADD45−/− CD4 T cells (77). Together, these results identify GADD45 proteins as important mediators of JNK and p38 signaling and of Th1 cytokine production. These studies have led to an interesting model for the regulation of p38 and JNK in Th differentiation: JNK and p38 are important for Th1 differentiation, and as a result of this process, there is a feedback amplification of signal strength for these two pathways. Rac2 and GADD45 proteins are induced, and as a consequence the signal transduction to these two MAP kinases is greatly enhanced. This greatly facilitates the effector cytokine production by Th1 cells.

MAP KINASES AND CYTOTOXICITY FUNCTION CD8+ cytotoxic T (Tc) cells are an important component of cellular immunity. Upon activation, Tc precursor cells proliferate with the help of IL-2, produce IFNγ and differentiate into cytotoxic T cells that secrete perforin and granzyme.

1 Feb 2002

10:8

66

AR

AR152-03.tex

DONG

¥

DAVIS

¥

AR152-03.SGM

LaTeX2e(2001/05/10)

P1: GSR

FLAVELL

Compared to CD4+ cells, there is relatively limited information on how Tc activation and function is regulated, and few studies have addressed the role of MAP kinases in these cells. One group studied the function of p38 MAP kinase in Tc cells. Merritt et al. found that p38 regulates IFN-γ production in CD8+ T cells as it does in CD4+ T cells (79), suggesting parallel pathways exist in CD4+ and CD8+ cells for IFNγ induction.

Annu. Rev. Immunol. 2002.20:55-72. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

MAP KINASES AND LYMPHOCYTE APOPTOSIS To restrict excessive inflammation and autoimmunity, the number of activated lymphocytes in our bodies is tightly regulated. This is achieved by activationinduced cell death mechanisms. MAP kinases are important for the regulation of cell death in a number of systems and are also implicated in regulation of cell survival and apoptosis in lymphocytes. JNK plays a role in thymocyte negative selection, probably through c-Jun phosphorylation (63, 70, 80, 81). In the periphery, JNK may also be involved in cell death regulation during activation-induced cell death. JNK1−/− cells exhibited greater proliferation, which was associated with reduced activation-induced cell death (61). This result suggests that JNK1 regulates apoptosis in T cells, as it does in other systems (61). However, how JNK1 works has not been established. JNK is involved in Fas-mediated cell death (82, 83), regulation of Fas ligand expression (84–87), and JNK may repress the function of antiapoptotic proteins bcl-2 and bclxl (88, 89), all of which could contribute to the above phenotype of JNK1−/− mice. p38 MAP kinase regulates activation-induced cell death of CD8+ cells. Activation of the p38 MAP kinase pathway in vivo induces apoptosis in CD8+ T cells, but not in CD4+ T cells, demonstrating that activation of the same specific signaling pathway can have a different outcome in the two T cell subsets (79). Induction of death in CD8+ T cells appears to be due to a reduction of Bcl2 levels selectively in CD8+ T cells, although this effect is mediated not by inhibition of Bcl-2 gene expression but most likely by increased degradation (79). Thus, activation of p38 MAP kinase results in increased production of IFN-γ by CD4+ Th1 cells and CD8+ T cells but also leads a decreased number of effector CD8+ T cells. Death/survival regulation in activated B lymphocytes is regulated by multiple cell-surface receptors. CD40 is an essential molecule for B cell survival and functional differentiation. Interestingly, CD40 preferentially activates JNK and p38 but not ERK (90). However, the role of JNK or other MAP kinases in regulation of in vivo B cell activation, function and cell death have not been defined genetically.

CONCLUSION AND FUTURE PERSPECTIVES MAP kinases are one of most ancient signal transduction pathways utilized by many systems and physiological processes. Their functions in immune responses are beginning to be revealed with help of kinase-specific inhibitors and mouse

1 Feb 2002

10:8

AR

AR152-03.tex

AR152-03.SGM

LaTeX2e(2001/05/10)

P1: GSR

Annu. Rev. Immunol. 2002.20:55-72. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

MAP KINASES

67

genetic manipulation. MAP kinases play important roles in all aspects of immune responses: from the innate to the adaptive immune system, from the initiation of immune responses to activation-induced cell death. We have also learned from these studies that the functions of given MAP kinases are subject to the context of different cell types and different immune responses. They have to collaborate with each other or with other signal transduction pathways. Together, these pathways draw a complex web of signal transduction in our immune cells. From what we summarized above, we can predict some future directions in the studies of MAP kinases: 1. Exploring the broader function of MAP kinases in different cell types of the immune system. For instance, the role of MAP kinases in dendritic cells, B cells, and CD8+ T cells are less characterized. These studies will be greatly facilitated by the availability of conditional knockout mouse models. 2. Defining the specific downstream targets of MAP kinases in a given stage and cell type of an immune response. MAP kinases are capable of phosphorylating multiple substrates. Which one mediates their function in a given reaction needs to be carefully characterized. Proteomics and targeted mutagensis of phosphorylation sites will assist these goals. 3. Understanding the signaling mechanisms of MAP kinases. MAP kinases are activated by many cell surface receptors. How these receptors transmit their signals to MAP kinase activation and how modules of MAP kinase pathways are assembled are still to be revealed. Studies of MKKK and scaffold proteins such as JIP will advance our basic knowledge of MAP kinase signal regulation. These results will no doubt advance our knowledge of the mechanisms of MAP kinase signaling in immune responses and may help development of therapeutic agents to selectively modulate MAP kinase activity to treat immune disorders. ACKNOWLEDGMENTS We thank our colleagues in our labs for scientific contribution and discussion, and Fran Manzo for secretarial help. C. Dong is an Arthritis Investigator award recipient. R.J. Davis and R.A. Flavell are investigators of the Howard Hughes Medical Institute. Visit the Annual Reviews home page at www.annualreviews.org

LITERATURE CITED 1. Schaeffe HJ, Weber MJ. 1999. Mitogenactivated protein kinases: specific messages from ubiquitous messengers. Mol. Cell. Biol, 19:2435–44 2. Han J, Ulevitch RJ. 1999. Emerging targets

for anti-inflammatory therapy. Nat. Cell. Biol. 1:E39–40 3. Davis RJ. 2000. Signal transduction by the JNK group of MAP kinases. Cell 103:239– 52

1 Feb 2002

10:8

Annu. Rev. Immunol. 2002.20:55-72. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

68

AR

AR152-03.tex

DONG

¥

DAVIS

¥

AR152-03.SGM

LaTeX2e(2001/05/10)

P1: GSR

FLAVELL

4. Chang L, Karin M. 2001. Mammalian MAP kinase signaling cascades. Nature 410:37–40 5. Sebzda E, Mariathasan S, Ohteki T, Jones R, Bachmann MF, Ohashi PS. 1999. Selection of the T cell repertoire. Annu. Rev. Immunol. 17:829–74 6. Rincon M. 2001. MAP-kinase signaling pathways in T cells. Curr. Opin. Immunol. 13:339–45 7. Rincon M, Flavell RA, Davis RJ. 2001. Signal transduction by MAP kinases in T lymphocytes. Oncogene 20:2490–97 8. Anderson KV. 2000. Toll signaling pathways in the innate immune response. Curr. Opin. Immunol. 12:13–19 9. Han ZS, Enslen H, Hu X, Meng X, Wu IH, Barrett T, Davis RJ, Ip YT. 1998. A conserved p38 mitogen-activated protein kinase pathway regulates Drosophila immunity gene expression. Mol. Cell. Biol. 18:3527–39 10. Sluss HK, Han Z, Barrett T, Davis RJ, Ip YT. 1996. A JNK signal transduction pathway that mediates morphogenesis and an immune response in Drosophila. Genes Dev. 10:2745–58 11. Medzhitov R, Janeway CA Jr. 1997. Innate immunity: the virtues of a nonclonal system of recognition. Cell 91:295–98 12. Hoshino K, Takeuchi O, Kawai T, Sanjo H, Ogawa T, Takeda Y, Takeda K, Akira S. 1999. Cutting edge: Toll-like receptor 4 (TLR4)-deficient mice are hyporesponsive to lipopolysaccharide: evidence for TLR4 as the Lps gene product. J. Immunol. 162:3749–52 13. Qureshi ST, Lariviere L, Leveque G, Clermont S, Moore KJ, Gros P, Malo D. 1999. Endotoxin-tolerant mice have mutations in Toll-like receptor 4 (Tlr4). J. Exp. Med. 189:615–25 14. Poltorak A, He X, Smirnova I, Liu MY, Huffel CV, Du X, Birdwell D, Alejos E, Silva M, Galanos C, Freudenberg M, Ricciardi-Castagnoli P, Layton B, Beutler B. 1998. Defective LPS signaling in C3H/HeJ and C57BL/10ScCr mice: mu-

15.

16.

17.

18.

19.

20.

21.

22.

23.

tations in Tlr4 gene. Science 282:2085– 88 Takeuchi O, Hoshino K, Kawai T, Sanjo H, Takada H, Ogawa T, Takeda K, Akira S. 1999. Differential roles of TLR2 and TLR4 in recognition of gram-negative and grampositive bacterial cell wall components. Immunity 11:443–51 Yang RB, Mark MR, Gray A, Huang A, Xie MH, Zhang M, Goddard A, Wood WI, Gurney AL, Godowski PJ. 1998. Toll-like receptor-2 mediates lipopolysaccharideinduced cellular signalling. Nature 395: 284–88 Hayashi F, Smith KD, Ozinsky A, Hown TR, Yi EC, Goodlett ER, Eng EK, Akira S, Underhill DM, Aderem A. 2001. The innate immune response to bacterial flagellin is mediated by Toll-like receptor 5. Nature 410:1099–1103 Hemmi H, Takeuchi O, Kawai T, Kaisho T, Sato S, Sanjo H, Matsumoto M, Hoshino K, Wagner H, Takeda K, Akira S. 2000. A Toll-like receptor recognizes bacterial DNA. Nature 408:740–45 Kawai T, Adachi O, Ogawa T, Takeda K, Akira S. 1999. Unresponsiveness of MyD88-deficient mice to endotoxin. Immunity 11:115–22 Medzhitov R, Preston-Hurlburt P, Kopp E, Stadlen A, Chen C, Ghosh S, Janeway CA Jr. 1998. MyD88 is an adaptor protein in the hToll/IL-1 receptor family signaling pathways. Mol. Cell. 2:253–58 Adachi O, Kawai T, Takeda K, Matsumoto M, Tsutsui H, Sakagami M, Nakanishi K, Akira S. 1998. Targeted disruption of the MyD88 gene results in loss of IL-1- and IL18-mediated function. Immunity 9:143–50 Burns K, Martinon F, Esslinger C, Pahl H, Schneider P, Bodmer JL, Di Marco F, French L, Tschopp J. 1998. MyD88, an adapter protein involved in interleukin-1 signaling. J. Biol. Chem. 273:12,203–9 Wesche H, Henzel WJ, Shillinglaw W, Li S, Cao Z. 1997. MyD88: an adapter that recruits IRAK to the IL-1 receptor complex. Immunity 7:837–47

1 Feb 2002

10:8

AR

AR152-03.tex

AR152-03.SGM

LaTeX2e(2001/05/10)

P1: GSR

Annu. Rev. Immunol. 2002.20:55-72. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

MAP KINASES 24. Muzio M, Ni J, Feng P, Dixit VM. 1997. IRAK (Pelle) family member IRAK-2 and MyD88 as proximal mediators of IL-1 signaling. Science 278:1612–15 25. Cao Z, Xiong J, Takeuchi M, Kurama T, Goeddel DV. 1996. TRAF6 is a signal transducer for interleukin-1. Nature 383:443–46 26. Kopp E, Medzhitov R, Carothers J, Xiao C, Douglas I, Janeway CA, Ghosh S. 1999. ECSIT is an evolutionarily conserved intermediate in the Toll/IL-1 signal transduction pathway. Genes Dev. 13:2059–71 27. Lomaga MA, Yeh WC, Sarosi I, Duncan GS, Furlonger C, Ho A, Morony S, Capparelli C, Van G, Kaufman S, van der Heiden A, Itie A, Wakeham A, Khoo W, Sasaki T, Cao Z, Penninger JM, Paige CJ, Lacey DL, Dunstan CR, Boyle WJ, Goeddel DV, Mak TW. 1999. TRAF6 deficiency results in osteopetrosis and defective interleukin-1, CD40, and LPS signaling. Genes Dev. 13:1015–24 28. Kanakaraj P, Schafer PH, Cavender DE, Wu Y, Ngo K, Grealish PF, Wadsworth SA, Peterson PA, Siekierka JJ, Harris CA, Fung-Leung WP. 1998. Interleukin (IL)-1 receptor-associated kinase (IRAK) requirement for optimal induction of multiple IL-1 signaling pathways and IL-6 production. J. Exp. Med. 187:2073–79 29. Lu HT, Yang DD, Wysk M, Gatti E, Mellman I, Davis RJ, Flavell RA. 1999. Defective IL-12 production in mitogen-activated protein (MAP) kinase kinase 3 (Mkk3)deficient mice. Embo J. 18:1845–57 30. Kotlyarov A, Neininger A, Schubert C, Eckert R, Birchmeier C, Volk HD, Gaestel M. 1999. MAPKAP kinase 2 is essential for LPS-induced TNF-alpha biosynthesis. Nat. Cell. Biol. 1:94–97 31. Kontoyiannis D, Pasparakis M, Pizarro TT, Cominelli F, Kollias G. 1999. Impaired on/off regulation of TNF biosynthesis in mice lacking TNF AU-rich elements: implications for joint and gut-associated immunopathologies. Immunity 10:387–98 32. Kontoyiannis D, Kotlyarov A, Carballo E, Alexopoulou L, Blackshear PJ, Gaestel

33.

34.

35.

36.

37.

38.

39.

40.

69

M, Davis R, Flavell R, Kollias G. 2001. Interleukin-10 targets p38 MAPK to modulate ARE-dependent TNF mRNA translation and limit intestinal pathology. EMBO J. 20:3760–70 Aoki M, Akiyama T, Miyoshi J, Toyoshima K. 1991. Identification and characterization of protein products of the cot oncogene with serine kinase activity. Oncogene 6:1515–19 Miyoshi J, Higashi T, Mukai H, Ohuchi T, Kakunaga T. 1991. Structure and transforming potential of the human cot oncogene encoding a putative protein kinase. Mol. Cell. Biol. 11:4088–96 Dumitru CD, Ceci JD, Tsatsanis C, Kontoyiannis D, Stamatakis K, Lin JH, Patriotis C, Jenkins NA, Copeland NG, Kollias G, Tsichlis PN. 2000. TNF-alpha induction by LPS is regulated posttranscriptionally via a Tpl2/ERK-dependent pathway. Cell 103:1071–83 Wysk M, Yang DD, Lu HT, Flavell RA, Davis RJ. 1999. Requirement of mitogenactivated protein kinase kinase 3 (MKK3) for tumor necrosis factor-induced cytokine expression. Proc. Natl. Acad. Sci. USA 96: 3763–68 Allen M, Svensson L, Roach M, Hambor J, McNeish J, Gabel CA. 2000. Deficiency of the stress kinase p38alpha results in embryonic lethality: characterization of the kinase dependence of stress responses of enzyme-deficient embryonic stem cells. J. Exp. Med. 191:859–70 Ganiatsas S, Kwee L, Fujiwara Y, Perkins A, Ikeda T, Labow MA, Zon LI. 1998. SEK1 deficiency reveals mitogen-activated protein kinase cascade crossregulation and leads to abnormal hepatogenesis. Proc. Natl. Acad. Sci. USA 95:6881–86 Chu WM, Ostertag D, Li ZW, Chang L, Chen Y, Hu Y, Williams B, Perrault J, Karin M. 1999. JNK2 and IKKbeta are required for activating the innate response to viral infection. Immunity 11:721–31 Han Z, Boyle DL, Chang L, Bennett B, Karin M, Yang L, Manning AM, Firestein

1 Feb 2002

10:8

70

41.

Annu. Rev. Immunol. 2002.20:55-72. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

42.

43.

44.

45. 46.

47.

48.

49.

50.

51.

AR

AR152-03.tex

DONG

¥

DAVIS

¥

AR152-03.SGM

LaTeX2e(2001/05/10)

P1: GSR

FLAVELL

GS. 2001. c-Jun N-terminal kinase is required for metalloproteinase expression and joint destruction in inflammatory arthritis. J. Clin. Invest. 108:73–81 Lee SY, Reichlin A, Santana A, Sokol KA, Nussenzweig MC, Choi Y. 1997. TRAF2 is essential for JNK but not NF-kappaB activation and regulates lymphocyte proliferation and survival. Immunity 7:703–13 Tournier C, Dong C, Turner TK, Jones SN, Flavell RA, Davis RJ. 2001. MKK7 is an essential component of the JNK signal transduction pathway activated by proinflammatory cytokines. Genes Dev. 15: 1419–26 Dong C, Flavell RA. 2000. Control of T helper cell differentiation—in search of master genes. Science’s STKE 49:pe1 Glimcher LH, Murphy KM. 2000. Lineage commitment in the immune system: the T helper lymphocyte grows up. Genes Dev. 14:1693–711 Paul WE, Seder RA. 1994. Lymphocyte response and cytokines. Cell 76:241–51 Kane LP, Lin J, Weiss A. 2000. Signal transduction by the TCR for antigen., Curr. Opin. Immunol. 12:242–49 Li W, Whaley CD, Mondino A, Mueller DL. 1996. Blocked signal transduction to the ERK and JNK protein kinases in anergic CD4+ T cells. Science 271:1272–76 Fields PE, Gajewski TF, Fitch FW. 1996. Blocked Ras activation in anergic CD4+ T cells. Science 271:1276–78 DeSilva DR, Jones EA, Favata MF, Jaffee BD, Magolda RL, Trzaskos JM, Scherle PA. 1998. Inhibition of mitogen-activated protein kinase kinase blocks T cell proliferation but does not induce or prevent anergy. J. Immunol. 160:4175–81 Pages G, Guerin S, Grall D, Bonino F, Smith A, Anjuere F, Auberger P, Pouyssegur J. 1999. Defective thymocyte maturation in p44 MAP kinase (Erk 1) knockout mice. Science 286:1374–77 Yamashita M, Kimura M, Kubo M, Shimizu C, Tada T, Perlmutter RM, Nakayama T. 1999. T cell antigen receptor-

52.

53.

54.

55.

56.

57.

58.

59.

60.

mediated activation of the Ras/mitogen- activated protein kinase pathway controls interleukin 4 receptor function and type-2 helper T cell differentiation. Proc. Natl. Acad. Sci. USA 96:1024–29 Rinc´on M, Enslen H, Raingeaud J, Recht M, Zapton T, Su MS-S, Penix LA, Davis RJ, Flavell RA. 1998. Interferon-γ expression by Th1 effector T cells mediated by the p38 MAP kinase signaling pathway. EMBO J. 17:2817–29 Yang J, Zhu H, Murphy TL, Ouyang W, Murphy KM. 2001. IL-18-stimulated GADD45β required in cytokine-induced, but not TCR-induced, IFNγ production. Nat. Immunol. 2:157–64 Zhang S, Kaplan MH. 2000. The p38 mitogen-activated protein kinase is required for IL-12-induced IFN-gamma expression. J. Immunol. 165:1374–80 Penix L, Weaver WM, Pang Y, Young HA, Wilson CB. 1993. Two essential regulatory elements in the human interferon gamma promoter confer activation specific expression in T cells. J. Exp. Med. 178:1483–96 Penix LA, Sweetser MT, Weaver WM, Hoey T, Hoeffler JP, Kerpolla TK, Wilson CB. 1996. The proximal regulatory element of the interferon-g promoter mediates selective expression in T cells. J. Biol. Chem. 271:31,964–72 Aune TM, Penix LA, Rinc´on M, Flavell RA. 1997. Memory and effector, but not naive T cells, express transcriptional activity under the control of two IFNg promoter elements. Mol. Cell. Biol. 17:199–204 Zhang F, Wang DZ, Boothby M, Penix LA, Flavell RA, Aune TM. 1998. Regulation of the activity of IFN-γ promoter elements during Th cell differentiation. J. Immunol. 161:6105–12 Reimold AM, Kim J, Finberg R, Glimcher LH. 2001. Decreased immediate inflammatory gene induction in activating transcription factor-2 mutant mice. Int. Immunol. 13:241–48 Weiss L, Whitmarsh AJ, Yang DD, Rincon M, Davis RJ, Flavell RA. 2000. Regulation

1 Feb 2002

10:8

AR

AR152-03.tex

AR152-03.SGM

LaTeX2e(2001/05/10)

P1: GSR

MAP KINASES

61.

Annu. Rev. Immunol. 2002.20:55-72. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

62.

63.

64.

65.

66.

67.

68.

69.

of c-Jun NH(2)-terminal kinase (Jnk) gene expression during T cell activation. J. Exp. Med. 191:139–46 Dong C, Yang DD, Wysk M, Whitmarsh AJ, Davis RJ, Flavell RA. 1998. Defective T cell differentiation in the absence of Jnk1. Science 282:2092–95 Rinc´on M, D´erijard B, Chow C-W, Davis RJ, Flavell RA. 1997b. Reprogramming the signaling requirement for AP-1 (activator protein-1) activation during differentiation of precursor CD4+ T cells to effector Th1 and Th2 cells. Genes Function 1:51– 68 Sabapathy K, Hu Y, Kallunki T, Schreiber M, David JP, Jochum W, Wagner EF, Karin M. 1999. JNK2 is required for efficient T-cell activation and apoptosis but not for normal lymphocyte development. Curr. Biol. 9:116–25 Yang DD, Conze D, Whitmarsh AJ, Barrett T, Davis RJ, Rincon M, Flavell RA. 1998. Differentiation of CD4+ T cells to Th1 cells requires MAP kinase JNK2. Immunity 9:575–85 Chen CY, Del Gatto-Konczak F, Wu Z, Karin M. 1998. Stabilization of interleukin2 mRNA by the c-Jun NH2-terminal kinase pathway. Science 280:1945–49 Chen CY, Gherzi R, Andersen JS, Gaietta G, Jurchott K, Royer HD, Mann M, Karin M. 2000. Nucleolin and YB-1 are required for JNK-mediated interleukin-2 mRNA stabilization during T-cell activation. Genes Dev. 14:1236–48 Constant SL, Dong C, Yang DD, Wysk M, Davis RJ, Flavell RA. 2000. JNK1 is required for T cell-mediated immunity against Leishmania major infection. J. Immunol. 165:2671–76 Lieberson R, Mowen KA, McBride KD, Leautaud V, Zhang X, Suh WK, Wu L, Glimcher LH. 2001. Tumor necrosis factor receptor-associated factor (TRAF)2 represses the T helper cell type 2 response through interaction with NFAT-interacting protein (NIP45). J. Exp. Med. 194:89–98 Chow C-W, Dong C, Flavell R, Davis

70.

71.

72.

73.

74.

75.

76.

77.

78.

71

R. 2000. JNK inhibits targeting of protein phosphatase calcineurin to NFATc1. Mol. Cell. Biol. 20:5227–34 Sabapathy K, Kallunki T, David JP, Graef I, Karin M. Wagner EF. 2001. c-Jun NH(2)-terminal kinase (JNK)1 and JNK2 have similar and stage-dependent roles in regulating T cell apoptosis and proliferation. J. Exp. Med. 193:317–28 Kuan CY, Yang DD, Samanta Roy DR, Davis RJ, Rakic P, Flavell RA. 1999. The Jnk1 and Jnk2 protein kinases are required for regional specific apoptosis during early brain development. Neuron 22:667–76 Sabapathy K, Jochum W, Hochedlinger K, Chang L, Karin M, Wagner EF. 1999. Defective neural tube morphogenesis and altered apoptosis in the absence of both JNK1 and JNK2. Mech. Dev. 89:115–24 Dong C, Yang D, Tournier C, Whitmarsh A, Xu J, Davis R, Flavell R. 2000. JNK is required for effector T-cell function but not for T-cell activation. Nature 405:91–94 Gong Q, Cheng AM, Akk AM, AlberolaIla J, Gong G, Pawson T, Chan AC. 2001. Disruption of T cell signaling networks and development by Grb2 haploid insufficiency. Nat. Immunol. 2:29–36 Li B, Yu H, Zheng W, Voll R, Na S, Roberts AW, Williams DA, Davis RJ, Ghosh S, Flavell RA. 2000. Role of the guanosine triphosphatase Rac2 in T helper 1 cell differentiation. Science 288:2219–22 Takekawa M, Saito H. 1998. A family of stress-inducible GADD45-like proteins mediate activation of the stress-responsive MTK1/MEKK4 MAPKKK. Cell 95:521– 30 Lu B, Yu H, Chow C, Li B, Zheng W, Davis RJ, Flavell RA. 2001. GADD45gamma mediates the activation of the p38 and JNK MAP kinase pathways and cytokine production in effector T(H)1 cells. Immunity 14:583–90 Hoffmeyer A, Piekorz R, Moriggl R, Ihle JN. 2001. Gadd45gamma is dispensable for normal mouse development and T-cell proliferation. Mol. Cell. Biol. 21:3137–43

1 Feb 2002

10:8

Annu. Rev. Immunol. 2002.20:55-72. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

72

AR

AR152-03.tex

DONG

¥

DAVIS

¥

AR152-03.SGM

LaTeX2e(2001/05/10)

P1: GSR

FLAVELL

79. Merritt C, Enslen H, Diehl N, Conze D, Davis RJ, Rincon M. 2000. Activation of p38 mitogen-activated protein kinase in vivo selectively induces apoptosis of CD8(+) but not CD4(+) T cells. Mol. Cell. Biol. 20:936–46 80. Rincon M, Whitmarsh A, Yang DD, Weiss L, Derijard B, Jayaraj P, Davis RJ, Flavell RA. 1998. The JNK pathway regulates the in vivo deletion of immature CD4(+)CD8(+) thymocytes. J. Exp. Med. 188:1817–30 81. Behrens A, Sabapathy K, Graef I, Cleary M, Crabtree GR, Wagner EF. 2001. Jun N-terminal kinase 2 modulates thymocyte apoptosis and T cell activation through c-Jun and nuclear factor of activated T cell (NF-AT). Proc. Natl. Acad. Sci. USA 98:1769–74 82. Yang X, Khosravi-Far R, Chang HY, Baltimore D. 1997. Daxx, a novel Fas-binding protein that activates JNK and apoptosis. Cell 89:1067–76 83. Wilson DJ, Fortner KA, Lynch DH, Mattingly RR, Macara IG, Posada JA, Budd RC. 1996. JNK, but not MAPK, activation is associated with Fas-mediated apoptosis in human T cells. Eur. J. Immunol. 26:989– 94 84. Le-Niculescu H, Bonfoco E, Kasuya Y, Claret FX, Green DR, Karin M. 1999. Withdrawal of survival factors results in activation of the JNK pathway in neuronal cells leading to Fas ligand induction and cell death. Mol. Cell. Biol. 19:751–63 85. Kasibhatla S, Brunner T, Genestier L, Echeverri F, Mahboubi A, Green DR.

86.

87.

88.

89.

90.

1998. DNA damaging agents induce expression of Fas ligand and subsequent apoptosis in T lymphocytes via the activation of NF-kappa B and AP-1. Mol. Cell. 1:543–51 Faris M, Latinis KM, Kempiak SJ, Koretzky GA, Nel A. 1998. Stress-induced Fas ligand expression in T cells is mediated through a MEK kinase 1-regulated response element in the Fas ligand promoter. Mol. Cell. Biol. 18:5414–24 Faris M, Kokot N, Latinis K, Kasibhatla S, Green DR, Koretzky GA, Nel A. 1998. The c-Jun N-terminal kinase cascade plays a role in stress-induced apoptosis in Jurkat cells by up-regulating Fas ligand expression. J. Immunol. 160:134–44 Yamamoto K, Ichijo H, Korsmeyer SJ. 1999. BCL-2 is phosphorylated and inactivated by an ASK1/Jun N-terminal protein kinase pathway normally activated at G(2)/M. Mol. Cell. Biol. 19:8469– 78 Kharbanda S, Saxena S, Yoshida K, Pandey P, Kaneki M, Wang Q, Cheng K, Chen YN, Campbell A, Sudha T, Yuan ZM, Narula J, Weichselbaum R, Nalin C, Kufe D. 2000. Translocation of SAPK/JNK to mitochondria and interaction with Bclx(L) in response to DNA damage. J. Biol. Chem. 275:322–27 Berberich I, Shu G, Siebelt F, Woodgett JR, Kyriakis JM, Clark EA. 1996. Cross-linking CD40 on B cells preferentially induces stress-activated protein kinases rather than mitogen-activated protein kinases. Embo J. 15:92–101

P1: FRK

February 1, 2002

12:47

Annual Reviews

AR152-FM

Annual Review of Immunology Volume 20, 2002

Annu. Rev. Immunol. 2002.20:55-72. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

CONTENTS FRONTISPIECE, Charles A. Janeway, Jr. A TRIP THROUGH MY LIFE WITH AN IMMUNOLOGICAL THEME, Charles A. Janeway, Jr.

xiv 1

THE B7 FAMILY OF LIGANDS AND ITS RECEPTORS: NEW PATHWAYS FOR COSTIMULATION AND INHIBITION OF IMMUNE RESPONSES, Beatriz M. Carreno and Mary Collins

29

MAP KINASES IN THE IMMUNE RESPONSE, Chen Dong, Roger J. Davis, and Richard A. Flavell

PROSPECTS FOR VACCINE PROTECTION AGAINST HIV-1 INFECTION AND AIDS, Norman L. Letvin, Dan H. Barouch, and David C. Montefiori T CELL RESPONSE IN EXPERIMENTAL AUTOIMMUNE ENCEPHALOMYELITIS (EAE): ROLE OF SELF AND CROSS-REACTIVE ANTIGENS IN SHAPING, TUNING, AND REGULATING THE AUTOPATHOGENIC T CELL REPERTOIRE, Vijay K. Kuchroo, Ana C. Anderson, Hanspeter Waldner, Markus Munder, Estelle Bettelli, and Lindsay B. Nicholson

55 73

101

NEUROENDOCRINE REGULATION OF IMMUNITY, Jeanette I. Webster, Leonardo Tonelli, and Esther M. Sternberg

125

MOLECULAR MECHANISM OF CLASS SWITCH RECOMBINATION: LINKAGE WITH SOMATIC HYPERMUTATION, Tasuku Honjo, Kazuo Kinoshita, and Masamichi Muramatsu

165

INNATE IMMUNE RECOGNITION, Charles A. Janeway, Jr. and Ruslan Medzhitov

KIR: DIVERSE, RAPIDLY EVOLVING RECEPTORS OF INNATE AND ADAPTIVE IMMUNITY, Carlos Vilches and Peter Parham ORIGINS AND FUNCTIONS OF B-1 CELLS WITH NOTES ON THE ROLE OF CD5, Robert Berland and Henry H. Wortis E PROTEIN FUNCTION IN LYMPHOCYTE DEVELOPMENT, Melanie W. Quong, William J. Romanow, and Cornelis Murre

197 217 253 301

LYMPHOCYTE-MEDIATED CYTOTOXICITY, John H. Russell and Timothy J. Ley

SIGNAL TRANSDUCTION MEDIATED BY THE T CELL ANTIGEN x

323

P1: FRK

February 1, 2002

12:47

Annual Reviews

AR152-FM

CONTENTS

xi

RECEPTOR: THE ROLE OF ADAPTER PROTEINS, Lawrence E. Samelson INTERACTION OF HEAT SHOCK PROTEINS WITH PEPTIDES AND ANTIGEN PRESENTING CELLS: CHAPERONING OF THE INNATE AND ADAPTIVE IMMUNE RESPONSES, Pramod Srivastava CHROMATIN STRUCTURE AND GENE REGULATION IN THE IMMUNE SYSTEM, Stephen T. Smale and Amanda G. Fisher PRODUCING NATURE’S GENE-CHIPS: THE GENERATION OF PEPTIDES FOR DISPLAY BY MHC CLASS I MOLECULES, Nilabh Shastri, Susan

371

Annu. Rev. Immunol. 2002.20:55-72. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

Schwab, and Thomas Serwold

395 427

463

THE IMMUNOLOGY OF MUCOSAL MODELS OF INFLAMMATION, Warren Strober, Ivan J. Fuss, and Richard S. Blumberg T CELL MEMORY, Jonathan Sprent and Charles D. Surh

GENETIC DISSECTION OF IMMUNITY TO MYCOBACTERIA: THE HUMAN MODEL, Jean-Laurent Casanova and Laurent Abel ANTIGEN PRESENTATION AND T CELL STIMULATION BY DENDRITIC CELLS, Pierre Guermonprez, Jenny Valladeau, Laurence Zitvogel, Clotilde Th´ery, and Sebastian Amigorena

495 551 581

621

NEGATIVE REGULATION OF IMMUNORECEPTOR SIGNALING, Andr´e Veillette, Sylvain Latour, and Dominique Davidson

669

CPG MOTIFS IN BACTERIAL DNA AND THEIR IMMUNE EFFECTS, Arthur M. Krieg

PROTEIN KINASE Cθ

709 IN

T CELL ACTIVATION, Noah Isakov and Amnon

Altman

RANK-L AND RANK: T CELLS, BONE LOSS, AND MAMMALIAN EVOLUTION, Lars E. Theill, William J. Boyle, and Josef M. Penninger PHAGOCYTOSIS OF MICROBES: COMPLEXITY IN ACTION, David M. Underhill and Adrian Ozinsky

761 795 825

STRUCTURE AND FUNCTION OF NATURAL KILLER CELL RECEPTORS: MULTIPLE MOLECULAR SOLUTIONS TO SELF, NONSELF DISCRIMINATION, Kannan Natarajan, Nazzareno Dimasi, Jian Wang, Roy A. Mariuzza, and David H. Margulies

853

INDEXES Subject Index Cumulative Index of Contributing Authors, Volumes 1–20 Cumulative Index of Chapter Titles, Volumes 1–20

ERRATA An online log of corrections to Annual Review of Immunology chapters (if any, 1997 to the present) may be found at http://immunol.annualreviews.org/

887 915 925

8 Feb 2002

9:36

AR

AR152-04.tex

AR152-04.SGM

LaTeX2e(2001/05/10) P1: GSR 10.1146/annurev.immunol.20.081501.094854

Annu. Rev. Immunol. 2002. 20:73–99 DOI: 10.1146/annurev.immunol.20.081501.094854 c 2002 by Annual Reviews. All rights reserved Copyright °

PROSPECTS FOR VACCINE PROTECTION AGAINST HIV-1 INFECTION AND AIDS

Annu. Rev. Immunol. 2002.20:73-99. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

Norman L. Letvin1, Dan H. Barouch1, and David C. Montefiori2 1

Harvard Medical School, Beth Israel Deaconess Medical Center, Boston, Massachusetts 02215; and 2Duke University Medical Center, Durham, North Carolina 27708; e-mail: [email protected]; dan [email protected]; [email protected]

■ Abstract The rapid and devastating spread of the AIDS epidemic in the developing world as well as the difficulties associated with delivering antiretroviral drugs in affected countries underscore the urgent need for the development of a safe and effective AIDS vaccine. In this review, we discuss recent advances in our understanding of the cellular and humoral immune responses to human immunodeficiency virus type 1 (HIV-1) infection. We then describe vaccine strategies that have been explored and discuss the evidence suggesting that cellular immune responses elicited by novel vaccine modalities may attenuate clinical disease caused by HIV-1.

INTRODUCTION Control of the worldwide AIDS epidemic requires an effective vaccine. However, the development of an AIDS vaccine has proven an enormous scientific challenge. Vaccine strategies effective in preventing infections with viruses such as smallpox, poliovirus, and hepatitis B have been ineffective in blocking AIDS virus transmission in animal models. Investigators therefore have studied the immune responses that contain HIV-1 replication in the setting of natural infection and have begun to develop vaccine strategies designed to generate these types of immune responses. In fact, accruing data have made a compelling argument that cell-mediated immunity plays a central role in containing HIV-1 replication. Moreover, recent studies in nonhuman primates suggest that vaccine-elicited memory cytotoxic T lymphocyte (CTL) populations can expand following an AIDS virus infection and can control viral replication. In the present article, recent progress in HIV-1 vaccine development is reviewed. The nonhuman primate models used for assessing HIV-1 vaccine strategies are described. Our current understanding of the roles of cellular and humoral immunity in the containment of HIV-1 replication are discussed, underscoring the contribution of CTL in controlling viral spread. Finally, recent studies are described, demonstrating that vaccine-elicited CTL populations can expand to limit AIDS 0732-0582/02/0407-0073$14.00

73

4 Feb 2002

11:19

74

AR

LETVIN

AR152-04.tex

¥

AR152-04.SGM

BAROUCH

¥

LaTeX2e(2001/05/10)

P1: GSR

MONTEFIORI

virus replication in nonhuman primates. These studies raise the possibility that currently available vaccine technologies may be able to contribute to the control of the HIV-1/AIDS epidemic.

Annu. Rev. Immunol. 2002.20:73-99. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

Nonhuman Primate Models for Evaluating Prototype HIV-1 Vaccines Animal models have played a central role in the elucidation of the immune responses to HIV-1 and in assessing strategies for vaccine protection against HIV-1 infection. Since HIV-1 cannot infect small laboratory animals or Old World monkeys, animal experimentation with this virus has been limited to studies in chimpanzees. The chimpanzee model for HIV-1 infection was for some time limited by the fact that patient isolates of HIV-1 replicate poorly and cause no clinical disease in chimpanzees. More recently, it has been shown that selected chimpanzeepassaged HIV-1 isolates can replicate to high levels and induce AIDS (1). However, this observation has not resulted in an increased use of the chimpanzee/HIV-1 model. Many investigators are unwilling for ethical reasons to infect chimpanzees with isolates of HIV-1 that cause AIDS. Furthermore, chimpanzees are available to investigators in only very small numbers and are extremely costly to use in experiments. For all of these reasons, investigators have turned to other AIDS animal models. HIV-1 is closely related to a number of primate lentiviruses that cause endemic infections in African nonhuman primate species. These viruses of nonhuman primates are known as simian immunodeficiency viruses (SIVs). Interestingly, these SIVs cause no disease in their natural host species (2). However, some SIV isolates can induce AIDS in Asian monkeys (3). Monkeys infected with these SIV isolates have provided a powerful model for exploring the immunopathogenesis of AIDS and assessing vaccine strategies for protecting against HIV-1 infection. Although the nucleotide sequences of the SIVs and HIV-1 share homologies and the interactions of these viruses with immune cells are similar, the envelopes of these viruses are quite divergent. This envelope sequence divergence has limited the utility of the SIV/macaque model for assessing envelope-based HIV-1 vaccine strategies. To overcome this limitation in the monkey AIDS model, chimeric viruses have been constructed that express HIV-1 envelopes on an SIV backbone (4–6). The in vivo passage of these chimeric simian-human immunodeficiency viruses (SHIVs) in monkeys has generated highly pathogenic viral isolates that cause rapid CD4+ T lymphocyte loss and death from AIDS (7). These viruses have been used in a number of recent trials of prototype HIV-1 vaccines in nonhuman primates.

Cellular Immune Responses Accumulating evidence over the past several years has confirmed the importance of cellular immune responses in controlling HIV-1 replication in humans and

4 Feb 2002

11:19

AR

AR152-04.tex

AR152-04.SGM

LaTeX2e(2001/05/10)

P1: GSR

Annu. Rev. Immunol. 2002.20:73-99. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

PROSPECTS FOR AIDS VACCINES

75

SIV replication in rhesus monkeys. The ability to elicit potent cellular immune responses has therefore become a priority for HIV-1 vaccine candidates. Early studies demonstrated that CD8+ T lymphocytes from HIV-1-infected individuals could suppress HIV-1 replication in autologous CD4+ T lymphocytes in vitro (8). Soon thereafter it was shown that HIV-1-specific CD8+ CTL were detectable in HIV-1-infected patients (9, 10). A temporal association was identified between the appearance of virus-specific CD8+ CTL in the peripheral blood and the decline of primary viremia in acutely infected patients (11–13). Since these CTL were observed well before serum neutralizing antibody activity could be detected, this observation suggested that CTL rather than neutralizing antibodies were responsible for the control of primary viremia. Similar observations were made in SIV-infected rhesus monkeys. During primary SIV infection of rhesus monkeys, a dramatic rise in CD8+ CTL with restricted Vβ repertoires occurred coincident with control of primary viremia (14, 15). The role of CTL has also been assessed in the control of chronic HIV-1 infection. HIV-1-infected patients who were clinical long-term nonprogressors had high levels of HIV-1-specific CTL (16). Moreover, in a prospective longitudinal study, HIV-1-infected patients with high virus-specific CTL activity had low viral loads, slow declines of peripheral blood CD4+ T lymphocyte counts, and a stable clinical status (17). These studies suggested that CTL responses controlled viral replication in chronically infected individuals. The recent development of quantitative techniques for the identification of epitope-specific CTL has allowed for a more detailed analysis of the magnitude and kinetics of the virus-specific CTL responses. Studies using fluorochromelabeled MHC class I tetramers demonstrated that HIV-1 tetramer binding CD8+ T lymphocytes were present in chronically HIV-1-infected individuals at remarkably high frequencies, often comprising over 1% of circulating CD8+ T cells (18). High levels of tetramer-positive CD8+ T lymphocytes were also found to coincide with the control of primary viremia in acutely infected individuals (19). Moreover, an inverse association between circulating tetramer-positive CD8+ T lymphocytes and viral load has been reported in some studies (20) but not others (R. Koup, unpublished). The discrepancy between the findings of these studies may reflect differences in levels of viral replication in the evaluated subjects. While potent CTL responses can act to contain levels of HIV-1 replication in infected individuals, it is also likely that persistent high levels of viral replication drive the expansion of CTL populations. Consistent with this notion, treatment of patients with highly active antiretroviral therapy resulted in a decline of tetramer-positive CD8+ T lymphocytes (21, 22). Tetramer staining has also been utilized to evaluate the role of CTL in SIVinfected rhesus monkeys. Dramatic expansions of SIV tetramer binding CD8+ T lymphocytes were observed during acute SIV infection, and high frequencies were maintained during chronic infection (23, 24). Direct evidence confirming the functional significance of these T cell responses has been obtained from studies showing that in vivo depletion of CD8+ T lymphocytes in SIV-infected monkeys

4 Feb 2002

11:19

Annu. Rev. Immunol. 2002.20:73-99. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

76

AR

LETVIN

AR152-04.tex

¥

AR152-04.SGM

BAROUCH

¥

LaTeX2e(2001/05/10)

P1: GSR

MONTEFIORI

resulted in a rapid and dramatic increase in viremia (25, 26). Moreover, vaccines that elicit potent CTL responses reduce viral replication following pathogenic viral challenges (27–32). Evidence from recent epidemiologic and clinical studies has further supported the functional importance of HIV-1-specific CTL responses in humans. Since MHC class I molecules present viral peptide fragments to CTL, investigators have reasoned that certain class I molecules may bind and display HIV-1 peptides more efficiently than other class I molecules. This might result in MHC class I–linked differences in viral containment. In fact, correlations have been demonstrated between certain HLA haplotypes and the rate of clinical disease progression in HIV1-infected individuals. HLA-B27 and HLA-B57 expression were associated with slow disease progression (33), whereas HLA-B*3501 and HLA-Cw*04 expression were associated with rapid disease progression (34, 35). Moreover, maximum heterozygosity of class I loci was associated with slow disease progression, whereas homozygosity was associated with rapid progression (34). These observations suggest that patients with a reduced breadth of potential CTL epitopes may have a reduced capacity for generating CTL responses that control viral replication. Other observations further strengthen the argument for the importance of cellular immune responses in controlling HIV-1 replication. Cohorts of seronegative Gambian and Kenyan commercial sex workers who were frequently exposed to HIV-1 yet remained uninfected had detectable HIV-1-specific CTL activity (36, 37). These findings raised the possibility that cellular immune responses may provide partial resistance to infection. The importance of virus-specific CTL in AIDS immunopathogenesis has also been highlighted by studies demonstrating that viral mutations that abrogate recognition by CTL have been associated with clinical progression to AIDS (38–41). A dramatic example of such viral escape from CTL recognition involved a patient who received an infusion of an expanded autologous Nef-specific CTL clone and subsequently developed an expanded population of nef-deleted virus associated with clinical disease progression (39). Experiments in SIV-infected rhesus monkeys have similarly described viral mutations that escape CTL recognition (42–44). Whereas the CTL responses observed in chronically HIV-1-infected individuals are potent and persistent, they ultimately fail to control virus replication in the majority of those who do not receive highly active antiretroviral therapy. Numerous mechanisms could account for this failure. The virus can escape from CTL recognition through mutation of epitopes (45). HIV-1 can cause downregulation of MHC class I expression through a process initiated by its Nef protein (46). CTL are also not capable of eliminating the reservoir of latently infected CD4+ T lymphocytes (47–49). In addition, recent data suggest that HIV-1-specific CTL may be functionally defective. HIV-1-specific CD8+ T cells have been reported to express a reduced level of perforin (50) and to have a CCR7− CD45RA− phenotype suggestive of incomplete maturation to terminally differentiated effector cells (51). However, the clinical relevance of this evidence for functionally impaired CTL is currently being debated (52).

4 Feb 2002

11:19

AR

AR152-04.tex

AR152-04.SGM

LaTeX2e(2001/05/10)

P1: GSR

Annu. Rev. Immunol. 2002.20:73-99. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

PROSPECTS FOR AIDS VACCINES

77

Lack of adequate CD4+ T cell help in HIV-1-infected individuals may be one of the main determinants of the functional impairment of CTL and the eventual failure of immune control of viremia. In lymphocytic choriomeningitis virus (LCMV) infection of mice lacking CD4+ T cell help, virus-specific CD8+ T cells were shown to lack antiviral effector function, suggesting a critical role for CD4+ T cells in maintaining optimal CTL function (53). Although HIV-1-infected individuals are characterized by potent virus-specific CD8+ CTL responses, their virus-specific CD4+ helper T lymphocyte responses are typically weak or undetectable (54). Only low levels of virus-specific CD4+ T cell responses in HIV-1-infected patients were detected by intracellular cytokine staining (55). In contrast, robust CD4+ T cell proliferative responses were demonstrated in clinical long-term nonprogressors, suggesting the importance of these responses in controlling viremia (56). Moreover, the early and aggressive treatment of acutely infected individuals rescues these virus-specific CD4+ T cell responses and allows selected individuals to maintain low viral loads even after the withdrawal of antiviral therapy (56, 57). These studies suggest that immunotherapeutic interventions that preserve HIV-1specific T cell help may improve immune control of viremia. The loss of HIV-1-specific CD4+ T cells may be central to the pathogenesis of AIDS. Deficient T cell help may undermine the functional integrity and survival of virus-specific CD8+ CTL, leading to eventual failure of immune control of viremia and clinical disease progression. It is therefore likely that effective HIV-1 vaccines will need to elicit potent virus-specific CD4+ as well as CD8+ T cell responses.

HUMORAL IMMUNE RESPONSES Although cellular immune responses play a central role in containing HIV-1 replication, humoral immune responses can also contribute to the control of the virus. Much of the evidence for this is derived from studies in nonhuman primates. Neutralizing monoclonal antibodies have recently been shown to confer passive protection to rhesus monkeys against challenge with a highly pathogenic SHIV that expresses the envelope glycoproteins of a primary HIV-1 isolate (58–60). Moreover, when neutralizing antibodies were elicited in monkeys immunized with Env glycoproteins, these monkeys were protected against SHIV challenges (61–63). These studies suggest that neutralizing antibodies may be capable of conferring protection against an AIDS virus infection. However, it has proven to be remarkably difficult to elicit antibodies that neutralize a diversity of primary virus isolates. This difficulty can be explained by the genetic variability and structural complexity of the HIV-1 envelope glycoproteins. The continuous generation of mutant viruses gives rise to HIV-1 variants that can escape humoral immune surveillance during the course of an infection (64). On a global scale, this genetic variation has likely given rise to an as yet poorly defined spectrum of neutralization serotypes that, with rare exceptions (65), are not predicted by genetic subtype (66, 67). In addition, some neutralization epitopes are

4 Feb 2002

11:19

Annu. Rev. Immunol. 2002.20:73-99. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

78

AR

LETVIN

AR152-04.tex

¥

AR152-04.SGM

BAROUCH

¥

LaTeX2e(2001/05/10)

P1: GSR

MONTEFIORI

poorly immunogenic (68, 69), while others appear to be masked by neighboring Nlinked glycans (70, 71) and tertiary folds in the viral envelope glycoproteins (72). Nonetheless, antibodies in the serum of a small subset of HIV-1-infected individuals are able to neutralize a broad spectrum of HIV-1 variants (16, 73). Moreover, three human monoclonal antibodies have been identified that have neutralizing activity for a diversity of HIV-1 isolates (74). The existence of these monoclonal antibodies provides what is perhaps the most compelling reason to believe it will be possible to elicit an antibody response that neutralizes a breadth of HIV-1 isolates with an immunogen or cassette of immunogens. Current efforts to develop immunogens for inducing neutralizing antibodies are based upon the presumed mechanism by which neutralization occurs and approaches to preserve or induce specific envelope configurations. Accumulating evidence supports a model of antibody-mediated HIV-1 neutralization that occurs at the interface of the virus and cell surface, presumably when a threshold number of envelope glycoprotein spikes are occupied by antibody (75). The envelope spikes of HIV-1 consist of surface gp120 molecules bound noncovalently to transmembrane gp41 molecules in a trimolecular complex of heterodimers. Entry is a multistep process that begins when gp120 engages CD4 on the surface of lymphocytes and macrophages (76). The thermodynamics of this interaction (77) are consistent with molecular and serologic observations (78, 79) as well as recent functional observations (80) suggesting that a stable conformational change in gp120 forms that exposes a highly conserved coreceptor binding site. The two most common coreceptors used by HIV-1 are the chemokine receptors CCR5 (R5 strains) and CXCR4 (X4 strains) (81, 82). By a mechanism that is not yet clear, coreceptor binding by gp120 triggers a change in the conformation of gp41 that exposes its hydrophobic fusion domain and permits insertion of the N-terminal peptide of gp41 into the cytoplasmic membrane of target cells (83–85). Further structural changes in this fusogenic form of gp41 then draw the virus and cell membranes into close proximity to complete the fusion process. These discrete steps in entry provide a number of opportunities for antibodies to neutralize HIV-1 (80, 86–88). However, all attempts to generate neutralizing antibodies have resulted in highly focused antibody responses that target a very narrow range of HIV-1 variants. The atomic structure of the gp120 core helps explain why it is difficult to generate cross-reactive neutralizing antibodies (89, 90). Conserved sequences comprising the docking sites for CD4 and coreceptor are in recessed pockets on the inner core of the gp120 molecule. These regions of gp120 are poorly accessible to most antibodies. Moreover, the outer domain of gp120 contains variable loops and sugar moieties that contribute to epitope masking. Although these observations are useful for understanding the antigenic properties of gp120, they must be viewed with an important caveat. This model is based on the atomic structure of a crystallized ternary complex of gp120 in association with the gp120-binding domain of CD4 and the Fab portion of a monoclonal antibody (17b) that is presumed to mimic coreceptor (91, 92). It will be important to generate

4 Feb 2002

11:19

AR

AR152-04.tex

AR152-04.SGM

LaTeX2e(2001/05/10)

P1: GSR

Annu. Rev. Immunol. 2002.20:73-99. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

PROSPECTS FOR AIDS VACCINES

79

structural information on Env in its native state, prior to receptor and coreceptor binding because this will be the usual target for antibodies. Information regarding neutralizing antibody epitopes on primary HIV-1 isolates has been limited to studies done using only three human monoclonal antibodies. Two of these monoclonal antibodies, IgG1b12 and 2G12, recognize gp120; the third, 2F5, recognizes gp41. The definition of the specificity of these three antibodies provides evidence that both Env subunits are potentially important immunogens. Only the 2F5 epitope has been mapped with certainty. It comprises a continuous region in the gp41 ectodomain that is mimicked antigenically by peptides containing the amino acid sequence ELDKWA (93). The IgG1b12 and 2G12 epitopes are more complex. The IgG1b12 epitope is in the CD4-binding domain of gp120 and is sensitive to mutations in V2 and C3 (94). The crystal structure of IgG1b12 has recently been solved, and when modeled with gp120, it revealed the presence of a protruding structure in the CDR3 region that penetrates the recessed CD4 binding pocket on gp120, possibly accounting for neutralizing activity (95). Consistent with this observation, recent evidence suggests that the hinge region of human IgG3 may also provide an advantage for HIV-1 neutralization (96). The 2G12 epitope differs from IgG1b12 in that it is comprised of residues in the C2-V4 region of gp120 and includes sites of N-glycosylation (97). As suggested by its three-dimensional position on the gp120 molecule, the 2G12 epitope may in fact be a glycan (90). Despite efforts by many investigators, the elicitation of antibodies with similar specificities using experimental immunogens has remained elusive.

HIV-1 VACCINE STRATEGIES: TRADITIONAL APPROACHES The first vaccine strategies that were rigorously evaluated as potential approaches for preventing HIV-1 infections were those that had previously proven effective in preventing human infections with other viral pathogens: live attenuated viruses, inactivated viruses, and subunit vaccines. Genetically altered, nonpathogenic live viruses have been used successfully to prevent polio, measles, and varicella. In 1992 the infection of monkeys with an SIV isolate made nonpathogenic by deletion of the gene encoding the accessory protein Nef was reported to protect monkeys from subsequent infection with wild-type pathogenic SIV (98). This finding generated a great deal of interest in the potential utility of live attenuated HIV-1 vaccines. However, a number of subsequent observations have raised serious questions as to the safety of this approach. A live attenuated SIV vaccine construct with a 12nucleotide deletion in the nef gene was shown to reconstruct a complete nef gene following infection of a monkey and regain pathogenic potential (99). Moreover, nef-deleted SIV isolates also induced AIDS in both neonatal monkeys and longterm infected adult monkeys (100, 101). Finally, a cohort of Australians infected by a single source of HIV-1-contaminated blood products was shown to harbor a virus isolate containing a significant mutation in the nef gene (102). While disease

4 Feb 2002

11:19

Annu. Rev. Immunol. 2002.20:73-99. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

80

AR

LETVIN

AR152-04.tex

¥

AR152-04.SGM

BAROUCH

¥

LaTeX2e(2001/05/10)

P1: GSR

MONTEFIORI

evolution may have been slow in these individuals, clinical manifestations of AIDS eventually became apparent (103, 104). These demonstrations of the pathogenic potential of live attenuated HIV-1 vaccine candidates have substantially dampened enthusiasm for this vaccine approach. No serious consideration is therefore currently being given to pursuing human efficacy trials with live attenuated HIV-1 vaccine prototypes. Inactivated virus vaccines have elicited immunity that prevents infection with poliovirus and influenza. While a report in 1989 described protection of monkeys from an SIV challenge using an inactivated virus vaccine, subsequent studies raised the possibility that this protection was an experimental artifact (105, 106). It is not surprising that useful protective immunity is not readily elicited by this strategy, since no virus-specific CTL responses can be induced by such an immunogen, and antiviral antibody responses generated by such an immunization are likely to be isolate-specific. Nevertheless, considerable effort has gone into the testing of an inactivated HIV-1 product in human volunteers (107, 108). This immunogen is for the most part depleted of envelope components of the virus during the process of its preparation. While it appears to be safe, no compelling evidence has been generated suggesting that this immunogen is efficacious. In view of the successful use of a recombinant protein as an immunogen for eliciting protective immunity against hepatitis B virus, considerable effort has been devoted to developing a recombinant protein vaccine for HIV-1. The monomeric recombinant envelope glycoprotein gp120 has been evaluated as an immunogen in nonhuman primate AIDS models and in human volunteers. Recombinant gp120immunized chimpanzees have been protected against challenge with a homologous HIV-1 (109). However, the ability to extrapolate from these studies to HIV-1infected humans has been called into question because the gp120 immunogen and challenge virus had identical envelope sequences, and the challenge viruses do not replicate to particularly high levels in chimpanzees. Therefore, many feel that these challenges do not represent a rigorous assessment of the protection that might be afforded by such a vaccine. Nevertheless, trials in human volunteers have assessed the safety and immunogenicity of this type of vaccine. These trials have shown that monomeric gp120 delivered as a protein with adjuvant does not elicit envelope-specific CTL responses. Moreover, the antibodies that have been elicited in these vaccinees do not neutralize primary patient isolates of HIV-1 (110). Finally, in limited studies of volunteers who were vaccinated with recombinant gp120 and subsequently became infected with HIV-1, no effect on the virologic or clinical outcome of these infections was shown (111). These findings convinced the National Institutes of Health not to proceed with efficacy trials of this type of immunogen. However, private vaccine manufacturers are currently carrying out efficacy trials of this type of vaccine in North America and Southeast Asia. Thus, the traditional strategies for creating vaccines have proven disappointing when applied to preventing infection with an AIDS virus. The reasons for the failures of these approaches are now clear. There is no evidence that one can uncouple the replication capacity of an AIDS virus and its pathogenicity. If the replication

4 Feb 2002

11:19

AR

AR152-04.tex

AR152-04.SGM

LaTeX2e(2001/05/10)

P1: GSR

Annu. Rev. Immunol. 2002.20:73-99. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

PROSPECTS FOR AIDS VACCINES

81

of a live, genetically altered immunodeficiency virus is sufficient to provide effective protection against infection by a wild-type immunodeficiency virus isolate, that genetically altered immunodeficiency virus will likely induce clinical disease itself. Moreover, the nature of the immune responses elicited by inactivated virus and recombinant monomeric protein immunogens does not appear adequate to contain the replication of a diverse population of HIV-1 isolates. Neither vaccine approach can elicit virus-specific CTL, and the antibodies generated through these approaches have been uniformly restricted in the diversity of viral isolates that they can neutralize. A growing recognition of the limitations of these traditional vaccine strategies has provided an impetus to investigators to explore novel vaccine strategies for preventing HIV-1 infections.

HIV-1 VACCINE STRATEGIES: NOVEL APPROACHES Considerable effort is currently being focused on the development and assessment of two novel strategies for vaccination: live vector-based approaches and plasmid DNA immunogens. Genes encoding proteins of HIV-1 can be inserted into the genomes of a variety of viruses and bacteria. When the resultant recombinant organisms infect a susceptible animal or human, immune responses are generated to both the parental organism and the products of the inserted HIV-1 genes. Since their antigens are processed by the immune system through the MHC class I pathway, the immune responses to the HIV-1 gene products include antibodies, helper T cells, and CTL. If infections with the parental virus or bacteria, the socalled vectors, do not lead to clinical disease, infections with the recombinant vectors are also well tolerated. To date, the live vectors that have been most extensively evaluated are the poxviruses. The prototype pox vector is vaccinia virus, the live attenuated virus that has been used in the successful worldwide campaign to eliminate smallpox. Immunization with recombinant vaccinia constructs elicits AIDS virus–specific cellular and humoral immunity in nonhuman primates (112, 113). Moreover, when combined with a protein boost, recombinant vaccinia immunization has protected monkeys from infection with some SIV isolates (114). There is, however, a reticence to move recombinant vaccinia viruses into human trials. It is well established that immunosuppressed individuals cannot contain vaccinia virus replication, and dissemination of the virus in immunocompromised individuals has been described (115). There is a fear that the administration of a recombinant vaccinia virus construct in a population of humans that includes a significant number of HIV1-infected, immunosuppressed individuals might result in significant vacciniaassociated morbidity and mortality in the vaccinees. Attention has therefore turned to a number of poxviruses that are attenuated for pathogenicity in humans. Modified vaccinia virus Ankara (MVA), generated by multiple passages of vaccinia virus in vitro, has a number of large genetic deletions that attenuate its replicative potential in primate cells (116). It is, therefore,

4 Feb 2002

11:19

Annu. Rev. Immunol. 2002.20:73-99. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

82

AR

LETVIN

AR152-04.tex

¥

AR152-04.SGM

BAROUCH

¥

LaTeX2e(2001/05/10)

P1: GSR

MONTEFIORI

likely to be nonpathogenic even in immunosuppressed humans. Recombinant MVA constructs have proven highly immunogenic in monkeys and have protected monkeys from SIV- and SHIV-induced disease (27, 31, 117, 118). Another multiple gene-deleted vaccinia vector, NYVAC, is also immunogenic in monkeys. While the absence of a proprietary position on MVA has slowed its development as an HIV-1 vaccine vector, recombinant MVA vaccines will soon enter safety and immunogenicity testing in the United States, Europe, and Africa. The best studied of the pox vectors are the avian poxviruses. While they do not undergo a complete replication cycle in human cells, they do initiate protein synthesis and therefore elicit immune responses in vaccinated individuals. In extensive safety and immunogenicity testing in human volunteers, recombinant canarypox constructs elicit low titer HIV-1-specific antibody responses in up to 70% of vaccinees. Moreover, at any single time of sampling vaccinees following immunization, 29% of individuals have detectable HIV-1-specific CTL responses demonstrable in their peripheral blood (119). Recombinant canarypox constructs also elicit CTL that recognize a diversity of HIV-1 clades (120). Whether this degree of immunogenicity warrants testing recombinant avian pox vectors for efficacy in humans is currently being discussed. Adenovirus is another vector being explored as a candidate AIDS vaccine. While the immunity elicited in preclinical studies with serotype 5 and 7 recombinant adenovirus vectors was disappointing, recent studies with gene-deleted adenoviruses originally developed for use in gene therapy have aroused considerable interest. These gene therapy vectors have been shown by a number of investigators to be highly immunogenic in small laboratory animals. Used to boost a plasmid DNA primed immune response, this type of vector was demonstrated to induce protective immunity in monkeys against infection with Ebola virus (121). Studies have recently been described showing impressive immunogenicity in monkeys with a recombinant serotype 5 adenovirus made replication incompetent by deletion of the E1 and E3 genes (32). Moreover, monkeys immunized with this vector were protected against SHIV-induced disease. These vectors have recently entered human trials to assess their safety and immunogenicity. Work with a number of other live viral vector systems is also in progress. Programs are pursuing the development of single strand RNA alpha virus vectors, including Venezuelan equine encephalitis virus and Semliki forest virus (122, 123). These vectors have already been shown to elicit immunodeficiency virus-specific CTL in small laboratory animals and nonhuman primates, and they should enter human safety trials in the near future. The parvovirus adeno-associated virus (AAV) has also been evaluated in nonhuman primates and has been shown to elicit potent SIV-specific CTL and reduce viral replication following SIV challenge (P. Johnson, unpublished). Because AAV integrates into the host genome, concerns of regulatory agencies may slow the progression of the testing of this vector system in humans. Finally, limited programs continue to pursue the use of recombinant bacteria as vector systems for an HIV-1 vaccine. The attenuated mycobacterium bacille

4 Feb 2002

11:19

AR

AR152-04.tex

AR152-04.SGM

LaTeX2e(2001/05/10)

P1: GSR

Annu. Rev. Immunol. 2002.20:73-99. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

PROSPECTS FOR AIDS VACCINES

83

Calmette-Guerin (BCG), engineered to carry HIV-1 and SIV genes, has been shown to elicit virus-specific CTL in monkeys (14). Attenuated enteric bacteria are also being explored as vector systems to carry HIV-1 DNA to the gastrointestinal mucosa with the intent of eliciting mucosal immune responses specific for those viral proteins (124). These programs, however, have to date garnered less interest than those pursuing live viral vectors. Perhaps the most novel technology being actively explored as a means of vaccinating to prevent infection with HIV-1 is the use of plasmid DNA inoculation (125). Following inoculation of small laboratory animals and monkeys, plasmid DNA vaccines express encoded viral proteins, and these proteins elicit both humoral and cellular immune responses. Considerable work has been done demonstrating that this level of immunity is sufficient to reduce viral replication in monkeys that are subsequently challenged with SIV and SHIV (28, 29). Moreover, monkeys that were immunized with DNA vaccines augmented by plasmid IL-2 constructs were protected against SHIV-induced disease (29). The ultimate utility of this vaccine modality in humans, however, remains an open question because preliminary data suggest that the immunogenicity of these plasmids may require relatively large doses. There is a growing concensus that plasmid DNA represents a particularly good priming immunogen. Therefore, considerable work is being done to explore the use of bimodal vaccine regimens in which the plasmid DNA is used to prime the immune response and a live recombinant vector is used to boost that immunity (30). No data have been generated indicating that any one of these novel vaccine modalities is superior to another in eliciting the immune responses needed to prevent HIV-1 infection in humans. Rather, there is a concensus that a number of these vaccine strategies should undergo extensive testing in human volunteers to assess their ultimate likelihood of utility. The immunogenicity of the live vectors may be limited in humans by pre-existing immunity to the vector organisms. The plasmid DNA constructs may not prove as immunogenic in humans as they have in mice and monkeys. A number of new approaches are being investigated for eliciting antibodies that may neutralize a variety of HIV-1 isolates. Some investigators have argued that an immunogen must preserve the native structure of oligomeric Env to be capable of inducing such an antibody response. This suggestion has arisen from evidence that antibody binding to monomeric gp120 does not predict the ability of that antibody to neutralize HIV-1 (126). It has, however, proven difficult to create a correctly folded and oligomerized Env immunogen. To avoid gp120-gp41 dissociation, early candidate immunogens were made that consisted of uncleaved Env gene products. These products were further modified to remove the transmembrane and cytoplasmic regions of gp41 to facilitate the purification of the secreted protein. Unfortunately, these uncleaved oligomers have proven no more effective as immunogens than monomeric gp120 (62, 127). Newer immunogens are in development that may more closely mimic the native Env structure (128, 129).

4 Feb 2002

11:19

Annu. Rev. Immunol. 2002.20:73-99. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

84

AR

LETVIN

AR152-04.tex

¥

AR152-04.SGM

BAROUCH

¥

LaTeX2e(2001/05/10)

P1: GSR

MONTEFIORI

Paradoxically, despite an abundance of native Env oligomer produced by replicating virus during the course of an infection, only a small fraction of HIV-1infected individuals have antibodies in their serum capable of neutralizing a broad spectrum of HIV-1 isolates. Some have suggested that unprocessed monomeric gp160 may act as an antigenic decoy to direct the B cell response away from potentially neutralizing epitopes on the native Env oligomer (130). Others have proposed that the V3 loop of gp120 is a decoy for deceptive imprinting (131). While these hypotheses are attractive, data to support them have not been forthcoming (132). Other investigators are exploring the ramifications of removing N-linked glycans and variable loop structures from Env that may be responsible for masking neutralization epitopes (71, 133). The hope is that by exposing these epitopes, reasonable titers of antibodies will be made that have a high affinity for Env. It is, however, unclear whether the antibodies elicited by such immunity will be capable of binding their cognate epitopes on wild-type virus particles and whether these antibodies actually have the capacity to neutralize a diversity of viral isolates. Approaches have also been taken to trigger and stabilize conformational intermediates of gp120 and gp41 that arise during the process of fusion. These preserved intermediates include cryptic epitopes on gp120 that become exposed upon CD4 ligation (134), and the prehairpin intermediate of gp41 that forms prior to the final steps in fusion (88). It has, however, been suggested that antibody might be sterically excluded from access to such triggered epitopes at the virus-cell interface. For example, the monoclonal antibodies 17b and CG10, which bind the coreceptor binding site of gp120, do not neutralize HIV-1 unless subinhibitory concentrations of recombinant soluble CD4 are added in vitro before the virus is incubated with cells (87). Another strategy suggested for inducing antibody responses that neutralize diverse viral isolates is to immunize with a mixture of envelope glycoproteins from HIV-1 strains that represent a spectrum of neutralization serotypes. It is not certain how many different Env immunogens would be needed to elicit such a response, nor is it certain what forms of Env protein should be employed as immunogens. Moreover, little information exists upon which to base a rationale for selection of those different Env proteins. Possible criteria might include genetic subtype, geographic distribution, biological phenotype (e.g., R5, X4, R5X4), and neutralization phenotype. Because of the time and cost associated with the manufacture of numerous recombinant Env proteins, little work has been done in this area (63). Finally, investigators have recently isolated antigens from phage-displayed peptide libraries that were recognized by either IgG1b12 (69) or neutralizing antibodies in the serum from HIV-1-infected individuals (135). These novel antigens might permit a detailed analysis of the structure of neutralization epitopes and might reveal strategies for enhancing the immunogenicity of such epitopes. None of these strategies has yet been successful in eliciting broadly neutralizing antibody responses in primates. However, if appropriate immunogens are eventually developed that are indeed capable of generating cross-reactive neutralizing

4 Feb 2002

11:19

AR

AR152-04.tex

AR152-04.SGM

LaTeX2e(2001/05/10)

P1: GSR

PROSPECTS FOR AIDS VACCINES

85

antibodies, suitable vectors may already exist for their delivery as part of a vaccine that is also capable of eliciting virus-specific CTL. A number of recombinant live vectors prime B cells for rapid secondary neutralizing antibody production following protein boosting or viral challenge (31, 123, 136, 137), suggesting that a variety of vectors may be effective in delivering Env immunogens.

Annu. Rev. Immunol. 2002.20:73-99. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

THERAPEUTIC IMMUNIZATION Currently available antiretroviral therapies for HIV-1-infected individuals have significant limitations. These limitations include cost, toxicities, the propensity to select for drug-resistant viral isolates, and the inability of these drugs to eliminate all reservoirs of virus in infected individuals. In light of accumulating evidence for the importance of cell-mediated immunity in containing HIV-1 replication and demonstrating development of novel vaccine strategies for eliciting cell-mediated immunity, some investigators have suggested that vaccines might be useful for augmenting virus-specific immune responses in infected individuals. Therapeutic vaccination and other immunologic interventions therefore may have the potential to be useful as adjunctive therapies. Perhaps the most convincing evidence that functional immunity to HIV-1 can be augmented in infected individuals has come from recent clinical studies involving structured or supervised treatment interruptions (STIs). The initiation of treatment in patients very early in the period of acute infection preserves HIV-1-specific CD4+ T cell responses in these individuals (56, 138). While stopping therapy in these patients led to an immediate and dramatic rise in the levels of replication virus, the circulating virus stimulated virus-specific immune responses. These immune responses were capable of controlling virus replication after the first interruption of therapy in a subset of individuals, and the majority of study subjects effectively controlled viral replication after subsequent STIs (57). Immune responses in these individuals following STIs were characterized by strong helper T cell responses and CTL responses focused on a limited number of epitopes. These highly focused CTL responses presumably reflected the fact that the virus population in these individuals was genetically quite homogeneous (139). The durability of this control of virus replication and the clinical ramifications of this degree of viral control, however, have not yet been determined. A similar phenomenon has also been reported in rhesus monkeys. In monkeys treated with antiretroviral drugs soon after SIV infection, successive STIs similarly resulted in augmented immune responses and good control of viral replication (140). These beneficial therapeutic effects of STIs, however, have only been achieved in the small fraction of patients treated very early in the period of acute infection. Treatment discontinuation or STIs in patients in whom antiretroviral drugs were initiated later during the course of infection have been significantly less successful (141, 142). In one study, eleven STIs in chronically HIV-1-infected individuals receiving highly active antiretroviral therapy resulted in transient increases in

4 Feb 2002

11:19

Annu. Rev. Immunol. 2002.20:73-99. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

86

AR

LETVIN

AR152-04.tex

¥

AR152-04.SGM

BAROUCH

¥

LaTeX2e(2001/05/10)

P1: GSR

MONTEFIORI

helper T cell responses, but no evidence of sustained control or viral replication (142). The use of therapeutic vaccines may provide a safer method for augmenting HIV-1-specific immune responses than the “autovaccination” that results from the burst of viral replication that occurs following treatment interruption. A preliminary study in SIV-infected rhesus monkeys treated during primary infection demonstrated that a recombinant poxvirus vaccine could augment cellular immune responses (143). A recent trial involving therapeutic vaccination of human volunteers using an HIV-1 immunogen, however, showed little effect (108). Another novel therapeutic vaccine approach that is currently being explored involves the injection of peptide-pulsed autologous dendritic cells (144, 145). It is hoped that vaccine-induced, enhanced immune responses may provide sustained immune control of viral replication and eventually allow the discontinuation of antiviral medications in infected individuals. Demonstrating clinically beneficial effects of therapeutic vaccination in the majority of chronically HIV-1infected patients, however, may prove to be considerably more difficult than doing so in acutely treated individuals.

VACCINE-ELICITED CELLULAR IMMUNE RESPONSES AND PREVENTION OF CLINICAL DISEASE It has been evident for many years to investigators assessing HIV-1 vaccine strategies in nonhuman primate models that vaccinated animals not protected from infection with a pathogenic immunodeficiency virus frequently survived for a prolonged period of time following viral challenge. It was, however, difficult to document this observation definitively until recently. A number of recent advances have facilitated a precise evaluation of this phenomenon. First, technologies have been developed that allow a precise quantitation of plasma viral load as well as vaccine-elicited CTL in monkeys (23, 146). The use of these technologies has allowed the assessment of correlations between vaccine-elicited immune responses and both virologic and clinical outcomes following immunodeficiency virus infections. Second, a correlation was conclusively demonstrated between viral load and rate of disease progression in previously unvaccinated SIV/SHIV-infected monkeys (117). This enabled investigators to hypothesize that set-point viral load may serve as a primary end point in vaccine challenge studies. Finally, immunodeficiency virus isolates have been developed that result in high and relatively uniform levels of viral replication in infected monkeys (7). This has allowed investigators to conduct vaccine studies with relatively small numbers of animals and to achieve statistically significant differences between experimental groups in virologic, immunologic, and clinical outcomes. Making use of these advances in technology and our understanding of AIDS pathogenesis, recent nonhuman primate studies have convincingly shown that immune responses elicited by a variety of vaccine modalities can alter the clinical

4 Feb 2002

11:19

AR

AR152-04.tex

AR152-04.SGM

LaTeX2e(2001/05/10)

P1: GSR

Annu. Rev. Immunol. 2002.20:73-99. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

PROSPECTS FOR AIDS VACCINES

87

course of an immunodeficiency virus infection (27–32). In these studies, monkeys have been vaccinated with immunogens such as plasmid DNA, cytokineaugmented plasmid DNA, recombinant MVA, recombinant gene–deleted adenovirus, and plasmid DNA followed by recombinant MVA, and then challenged with the highly pathogenic SHIV-89.6P or SIVsmE660. These immunizations led to the induction of virus-specific CTL, and the frequency of vaccine-elicited virusspecific CTL correlated with the containment of viral replication following the challenge. Moreover, these studies also demonstrated that the level of detectable plasma virus correlated with the rapidity of disease evolution. Therefore, potent vaccine-elicited CTL diminished viral loads and, accordingly, slowed clinical disease progression following viral challenge. The implications of these findings for HIV-1 vaccine development in humans are substantial. Clearly, the ideal AIDS vaccine remains one that will induce sterilizing immunity, preventing infection following exposure to the virus. It is likely that such sterilizing immunity will only be elicited once we understand how to induce an antibody response that can neutralize a diversity of HIV-1 isolates. In fact, it is likely that it will be necessary to elicit both CTL and a broadly neutralizing antibody response to achieve protection against infection. Considerable efforts continue to be devoted to solving the problem of eliciting such a neutralizing antibody response. However, the data generated in the recent series of nonhuman primate studies (27–32) suggest that we may be in a position at this time to provide meaningful protection against the clinical sequelae of HIV-1 infection even if we are unable to provide protection against infection with the virus. At least some of the vaccine technologies that have proven effective in eliciting immunodeficiency virus-specific CTL responses in monkeys should elicit such responses in humans. If these nonhuman primate studies are predictive of what will be seen in humans, individuals who have vaccine-elicited CTL responses and are subsequently infected with HIV-1 may contain the virus more effectively than unvaccinated individuals. This may result in two beneficial sequelae. First, these individuals may have a more prolonged survival following infection than do unvaccinated individuals (147). Second, with lower viral burdens in their secretions, these individuals may have a lower likelihood of transmitting virus to uninfected individuals. Recent studies in human populations have shown that the risk of transmitting HIV-1 from an infected to uninfected individual correlates with the plasma viral load of the infected subject; individuals with high viral loads are more likely to transmit the virus than individuals with low viral loads (148). It is therefore possible that the transmission of HIV-1 may be substantially slowed in a population that is vaccinated with an immunogen that elicits CTL. The coming years should be exciting ones in the arena of AIDS vaccine development. Investigators will be redoubling their efforts to develop strategies for eliciting antibodies that can neutralize a diversity of HIV-1 isolates. Work will also be devoted to generating more effective regimens for inducing HIV-1-specific CTL responses. Most importantly, safety, immunogenicity, and finally efficacy

4 Feb 2002

11:19

88

AR

LETVIN

AR152-04.tex

¥

AR152-04.SGM

BAROUCH

¥

LaTeX2e(2001/05/10)

P1: GSR

MONTEFIORI

trials will be performed to assess a number of strategies for eliciting HIV-1-specific CTL in humans, with the hope that some of these vaccines may attenuate disease in individuals who become infected with this virus. Increasing the duration and improving the quality of life of individuals who become infected with HIV-1 through prior vaccination may actually be an achievable goal today.

Annu. Rev. Immunol. 2002.20:73-99. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

CONCLUSIONS Recent advances in our understanding of the immunopathogenesis of HIV-1 infection are currently being harnessed in the development of a new generation of candidate AIDS vaccines. Passive infusions of neutralizing antibodies in rhesus monkeys have demonstrated that complete protection against pathogenic AIDS virus challenges is possible. However, vaccine approaches for eliciting antibodies that neutralize a diversity of HIV-1 isolates remain elusive. A number of vaccine strategies that elicit potent virus-specific cellular immune responses have already been developed. Immune responses elicited by these vaccines control viral replication and prevent clinical AIDS in rhesus monkeys following pathogenic viral challenges. These studies suggest that currently available vaccine technologies may be capable of eliciting immune responses that can slow clinical disease progression and reduce transmission rates by lowering viral loads even if they fail to provide sterilizing immunity in human populations. However, these findings in monkey models may not accurately predict the utility of these vaccines in humans. Human trials evaluating the safety, immunogenicity, and efficacy of these promising vaccine candidates are now being pursued. ACKNOWLEDGMENTS Visit the Annual Reviews home page at www.annualreviews.org

LITERATURE CITED 1. Novembre FJ, Saucier M, Anderson DC, Klumpp SA, O’Neil SP, Brown CR, Hart CE, Guenthner PC, Swenson RB, McClure HM. 1997. Development of AIDS in a chimpanzee infected with human immunodeficiency virus type 1. J. Virol. 71:4086– 91 2. Hirsch VM, Johnson PR. 1994. Pathogenic diversity of simian immunodeficiency viruses. Virus Res. 32:183–203 3. Letvin NL, King NW. 1990. Immunologic and pathologic manifestations of the infec-

tion of rhesus monkeys with simian immunodeficiency virus of macaques. J. Acquir. Immune Defic. Synd. 3:1023–40 4. Li JT, Lord CI, Haseltine W, Letvin NL, Sodroski JG. 1992. Infection of cynomolgus monkeys with a chimeric HIV-1/ SIVmac virus that expresses the HIV-1 envelope glycoproteins. J. Acquir. Immune Defic. Syndr. 5:639–46 5. Li JT, Halloran M, Lord CI, Watson A, Ranchalis J, Fung M, Letvin NL, Sodroski JG. 1995. Persistent infection of macaques

4 Feb 2002

11:19

AR

AR152-04.tex

AR152-04.SGM

LaTeX2e(2001/05/10)

P1: GSR

PROSPECTS FOR AIDS VACCINES

6.

Annu. Rev. Immunol. 2002.20:73-99. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

7.

8.

9.

10.

11.

12.

13.

with simian-human immunodeficiency viruses. J. Virol. 69:7061–67 Luciw PA, Pratt-Lowe E, Shaw KE, Levy JA, Cheng-Mayer C. 1995. Persistent infection of rhesus macaques with T-cellline-tropic and macrophage-tropic clones of simian/human immunodeficiency viruses (SHIV). Proc. Natl. Acad. Sci. USA 92:7490–94 Reimann KA, Li JT, Veazey R, Halloran M, Park I-W, Karlsson GB, Sodroski J, Letvin NL. 1996. A chimeric simian/ human immunodeficiency virus expressing a primary patient human immunodeficiency virus type 1 isolate env causes an AIDS-like disease after in vivo passage in rhesus monkeys. J. Virol. 70:6922–28 Walker CM, Moody DJ, Stites DP, Levy JA. 1986. CD8+ lymphocytes can control HIV infection in vitro by suppressing virus replication. Science 234:1563–66 Walker BD, Chakrabarti S, Moss B, Paradis TJ, Flynn T, Durno AG, Blumberg RS, Kaplan JC, Hirsch MS, Schooley RT. 1987. HIV-specific cytotoxic T lymphocytes in seropositive individuals. Nature 328:345–48 Plata F, Autran B, Martins LP, WainHobson S, Raphael M, Mayaud C, Denis M, Guillon JM, Debre P. 1987. AIDS virus-specific cytotoxic T lymphocytes in lung disorders. Nature 328:348–51 Koup RA, Safrit JT, Cao Y, Andrews CA, McLeod G, Borkowsky W, Farthing C, Ho DD. 1994. Temporal association of cellular immune responses with the initial control of viremia in primary human immunodeficiency virus type 1 syndrome. J. Virol. 68:4650–55 Borrow P, Lewicki H, Hahn BH, Shaw GM, Oldstone MBA. 1994. Virus-specific CD8+ cytolytic T-lymphocyte activity associated with control of viremia in primary human immunodeficiency virus type 1 infection. J. Virol. 68:6103–10 Pantaleo G, Demarest JF, Soudeyns H, Graziosi C, Denis F, Adelsberger JW, Borrow P, Saag MS, Shaw GM, Sekaly RP,

14.

15.

16.

17.

18.

19.

20.

89

Fauci AS. 1994. Major expansion of CD8+ T cells with a predominant Vβ usage during the primary immune response to HIV. Nature 370:463–67 Yasutomi Y, Reimann KA, Lord CI, Miller MD, Letvin NL. 1993. Simian immunodeficiency virus specific CD8+ lymphocyte response in acutely infected rhesus monkeys. J. Virol. 67:1707–11 Chen ZW, Kou ZC, Lekutis C, Shen L, Zhou D, Halloran M, Li J, Sodroski J, Lee-Parritz D, Letvin NL. 1995. T cell receptor Vβ repertoire in an acute infection of rhesus monkeys with simian immunodeficiency viruses and a chimeric simianhuman immunodeficiency virus. J. Exp. Med. 182:21–31 Cao Y, Qin L, Zhang L, Safrit J, Ho DD. 1995. Virologic and immunologic characterization of long-term survivors of human immunodeficiency virus type 1 infection. New Engl. J. Med. 332:201–8 Musey L, Hughes J, Schacker T, Shea T, Corey L, McElrath MJ. 1997. Cytotoxic-Tcell responses, viral load, and disease progression in early human immunodeficiency virus type 1 infection. New Engl. J. Med. 337:1267–74 Altman JD, Moss PAH, Goulder PJR, Barouch DH, McHeyzer-Williams MG, Bell JI, McMichael AJ, Davis MM. 1996. Phenotypic analysis of antigen-specific T lymphocytes. Science 274:94–96 Wilson JD, Ogg GS, Allen RL, Davis C, Shaunak S, Downie J, Dyer W, Workman C, Sullivan S, McMichael AJ, RowlandJones SL. 2000. Direct visualization of HIV-1-specific cytotoxic T lymphocytes during primary infection. AIDS 14:225– 33 Ogg GS, Jin X, Bonhoeffer S, Dunbar PR, Nowak MA, Monard S, Segal JP, Cao Y, Rowland-Jones SL, Cerundolo V, Hurley A, Markowitz M, Ho DD, Nixon DF, McMichael AJ. 1998. Quantitation of HIV-1–specific cytotoxic T lymphocytes and plasma load of viral RNA. Science 279:2103–6

4 Feb 2002

11:19

Annu. Rev. Immunol. 2002.20:73-99. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

90

AR

LETVIN

AR152-04.tex

¥

AR152-04.SGM

BAROUCH

¥

LaTeX2e(2001/05/10)

P1: GSR

MONTEFIORI

21. Ogg GS, Jin X, Bonhoeffer S, Moss P, Nowak MA, Monard S, Segal JP, Cao Y, Rowland-Jones SL, Hurley A, Markowitz M, Ho DD, McMichael AJ, Nixon DF. 1999. Decay kinetics of human immunodeficiency virus-specific effector cytotoxic T lymphocytes after combination antiretroviral therapy. J. Virol. 73:797–800 22. Kalams SA, Goulder PJ, Shea AK, Jones NG, Trocha AK, Ogg GS, Walker BD. 1999. Levels of human immunodeficiency virus type 1-specific cytotoxic T-lymphocyte effector and memory responses decline after suppression of viremia with highly active antiretroviral therapy. J. Virol. 73:6721–28 23. Kuroda MJ, Schmitz JE, Barouch DH, Criau A, Allen TM, Sette A, Watkins DI, Forman MA, Letvin NL. 1998. Analysis of Gag-specific cytotoxic T lymphocytes in simian immunodeficiency virusinfected rhesus monkeys by cell staining with a tetrameric major histocompatibility complex class I-peptide complex. J. Exp. Med. 187:1373–81 24. Kuroda MJ, Schmitz JE, Charini WA, Nickerson CE, Lifton MA, Lord CI, Forman MA, Letvin NL. 1999. Emergence of CTL coincides with clearance of virus during primary simian immunodeficiency virus infection in rhesus monkeys. J. Immunol. 162:5127–33 25. Schmitz JE, Kuroda MJ, Santra S, Sasseville VG, Simon MA, Lifton MA, Racz P, Tenner-Racz K, Dalesandro M, Scallon BJ, Ghrayeb J, Forman MA, Montefiori DC, Rieber EP, Letvin NL, Reimann KA. 1999. Control of viremia in simian immunodeficiency virus infection by CD8+ lymphocytes. Science 283:857–60 26. Jin X, Bauer DE, Tuttleton SE, Lewin S, Gettie A, Blanchard J, Irwin CE, Safrit JT, Mittler J, Weinberger L, Kostrikis LG, Zhang L, Perelson AS, Ho DD. 1999. Dramatic rise in plasma viremia after CD8+ T cell depletion in simian immunodeficiency virus-infected macaques. J. Exp. Med. 189:991–98

27. Seth A, Ourmanov I, Schmitz JE, Kuroda MJ, Lifton MA, Nickerson CE, Wyatt L, Carroll M, Moss B, Venzon D, Letvin NL, Hirsch VM. 2000. Immunization with a modified vaccinia virus expressing simian immunodeficiency virus (SIV) Gag-Pol primes for an anamnestic Gag-specific cytotoxic T-lymphocyte response and is associated with reduction of viremia after SIV challenge. J. Virol. 74:2502–9 28. Egan MA, Charini WA, Kuroda MJ, Voss G, Schmitz JE, Racz P, Tenner-Racz K, Manson K, Wyand M, Lifton MA, Nickerson CE, Fu T-M, Shiver JW, Letvin NL. 2000. Simian immunodeficiency virus (SIV) gag DNA-vaccinated rhesus monkeys develop secondary cytotoxic T-lymphocyte responses and control viral replication after pathogenic SIV infection. J. Virol. 74:7485–95 29. Barouch DH, Santra S, Schmitz JE, Kuroda MJ, Fu T-M, Wagner W, Bilska M, Craiu A, Zheng XX, Krivulka GR, Beaudry K, Lifton MA, Nickerson CE, Trigona WL, Punt K, Freed DC, Guan L, Dubey S, Casimiro D, Simon A, Davies M-E, Chastain M, Strom TB, Gelman RS, Montefiori DC, Lewis MG, Emini EA, Shiver JW, Letvin NL. 2000. Control of viremia and prevention of clinical AIDS in rhesus monkeys by cytokine-augmented DNA vaccination. Science 290:486–92 30. Amara RR, Villinger F, Altman JD, Lydy SL, O’Neil SP, Staprans SI, Montefiori DC, Xu Y, Herndon JG, Wyatt LS, Candido MA, Kozyr NL, Earl PL, Smith JM, Ma H-K, Grimm BD, Hulsey ML, McClure HM, McNicholl JM, Moss B, Robinson HL. 2001. Control of a mucosal challenge and prevention of clinical AIDS in rhesus monkeys by a multiprotein DNA/ MVA vaccine. Science 292:69–74 31. Barouch DH, Santra S, Kuroda MJ, Schmitz JE, Plishka R, Buckler-White A, Gaitan AE, Zin R, Nam J-H, Wyatt LS, Lifton MA, Nickerson CE, Moss B, Montefiori DC, Hirsch VM, Letvin NL. 2001. Reduction of simian-human immunodeficiency

4 Feb 2002

11:19

AR

AR152-04.tex

AR152-04.SGM

LaTeX2e(2001/05/10)

P1: GSR

PROSPECTS FOR AIDS VACCINES

32.

Annu. Rev. Immunol. 2002.20:73-99. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

33.

34.

35.

36.

37.

38.

virus 89.6P viremia in rhesus monkeys by recombinant modified vaccinia virus Ankara vaccination. J. Virol. 75:5151– 58 Shiver JW. 2001. Comparative studies of severla HIV vaccine approaches. AIDS Vaccines in the New Millenium. Keystone Symposia, Keystone, CO Kaslow RA, Carrington M, Apple R, Park L, Munoz A, Saah AJ, Goedert JJ, Winkler C, O’Brien SJ, Rinaldo C, Detels R, Blattner W, Phair J, Erlich H, Mann DL. 1996. Influence of combinations of human major histocompatibility complex genes on the course of HIV-1 infection. Nat. Med. 2:405–11 Carrington M, Nelson GW, Martin MP, Kissner T, Vlahov D, Goedert JJ, Kaslow R, Buchbinder S, Hoots K, O’Brien SJ. 1999. HLA and HIV-1: heterozygote advantage and B*35-Cw*04 disadvantage. Science 283:1748–52 Gao X, Nelson GW, Karacki P, Martin MP, Phair J, Kaslow R, Goedert JJ, Buchbinder S, Hoots K, Vlahov D, O’Brien SJ, Carrington M. 2001. Effect of a single amino acid change in MHC class I molecules on the rate of progression to AIDS. New Engl. J. Med. 344:1668–75 Rowland-Jones S, Sutton J, Ariyoshi K, Dong T, Gotch F, McAdam S, Whitby D, Sabally S, Gallimore A, Corrah T, Takiguchi M, Schultz T, McMichael A, Whittle H. 1995. HIV-specific cytotoxic T-cells in HIV-exposed but uninfected Gambian women. Nat. Med. 1:59–64 Kaul R, Plummer FA, Kimani J, Dong T, Kiama P, Rostron T, Njagi E, MacDonald KS, Bwayo JJ, McMichael AJ, RowlandJones SL. 2000. HIV-1-specific mucosal CD8+ lymphocyte responses in the cervix of HIV-1-resistant prostitutes in Nairobi. J. Immunol. 164:1602–11 Phillips RE, Rowland-Jones S, Nixon DF, Gotch FM, Edwards JP, Ogunlesi AO, Elvin JG, Rothbard JA, Bangham CRM, Rizza CR, McMichael AJ. 1991. Human immunodeficiency virus genetic variation

39.

40.

41.

42.

43.

44.

91

that can escape cytotoxic T cell recognition. Nature 354:453–59 Koenig S, Conley AJ, Brewah YA, Jones GM, Leath S, Boots LJ, Davey V, Pantaleo G, Demarest JF, Carter C. 1995. Transfer of HIV-1-specific cytotoxic T lymphocytes to an AIDS patient leads to selection for mutant HIV variants and subsequent disease progression. Nat. Med. 1:330–36 Goulder PJ, Tang Y, Brander C, Betts MR, Altfeld M, Annamalai K, Trocha A, He S, Rosenberg ES, Ogg G, O’Callaghan CA, Kalams SA, McKinney Jr RE, Mayer K, Koup RA, Pelton SI Burchett SK, McIntosh K, Walker BD. 2000. Functionally inert HIV-specific cytotoxic T lymphocytes do not play a major role in chronically infected adults and children. J. Exp. Med. 192:1819–32 Price DA, Goulder PJ, Klenerman P, Sewell AK, Easterbrook PJ, Troop M, Bangham CR, Phillips RE. 1997. Positive selection of HIV-1 cytotoxic T lymphocyte escape variants during primary infection. Proc. Natl. Acad. Sci. USA 94:1890–95 Evans DT, O’Connor DH, Jing P, Dzuris JL, Sidney J, da Silva J, Allen TM, Horton H, Venham JE, Rudersdorf RA, Vogel T, Pauza CD, Bontrop RE, DeMars R, Sette A, Hughes AL, Watkins DI. 1999. Virusspecific cytotoxic T-lymphocyte responses select for amino-acid variation in simian immunodeficiency virus Env and Nef. Nat. Med. 5:1270–76 Allen TM, O’Connor DH, Jing P, Dzuris JL, Mothe BR, Vogel TU, Dunphy E, Liebl ME, Emerson C, Wilson N, Kunstman KJ, Wang X, Allison DB, Hughes AL, Desrosiers RC, Altman JD, Wolinsky SM, Sette A, Watkins DI. 2000. Tat-specific cytotoxic T lymphocytes select for SIV escape variants during resolution of primary viremia. Nature 407:386–90 Chen ZW, Craiu A, Shen L, Kuroda MJ, Iroku UC, Watkins DI, Voss G, Letvin NL. 2000. Simian immunodeficiency virus evades a dominant epitope-specific cytotoxic T lymphocyte response through a mutation

4 Feb 2002

11:19

92

45.

Annu. Rev. Immunol. 2002.20:73-99. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

46.

47.

48.

49.

50.

51.

52.

AR

LETVIN

AR152-04.tex

¥

AR152-04.SGM

BAROUCH

¥

LaTeX2e(2001/05/10)

P1: GSR

MONTEFIORI

resulting in the accelerated dissociation of viral peptide and MHC class I. J. Immunol. 164:6474–79 McMichael AJ, Phillips RE. 1997. Escape of human immunodeficiency virus from immune control. Annu. Rev. Immunol. 15:271–96 Collins KL, Chen BK, Kalams SA, Walker BD, Baltimore D. 1998. Nef protein protects infected primary cells against killing by cytotoxic T lymphocytes. Nature 391:397–401 Wong JK, Hezareh M, Gunthard HF, Havlir DV, Ignacio CC, Spina CA, Richman DD. 1997. Recovery of replicationcompetent HIV despite prolonged suppression of plasma viremia. Science 278:1291– 95 Finzi D, Hermankova M, Pierson T, Carruth LM, Buck C, Chaisson RE, Quinn TC, Chadwick K, Margolick J, Brookmeyer R, Gallant J, Markowitz M, Ho DD, Richman DD, Siliciano RF. 1997. Identification of a reservoir for HIV-1 in patients on highly active antiretroviral therapy. Science 278:1295–1300 Chun T-W, Stuyver L, Mizell SB, Ehler LA, Mican JM, Baseler M, Lloyd AL, Nowak MA, Fauci AS. 1997. Presence of an inducible HIV-1 latent reservoir during highly active antiretroviral therapy. Proc. Natl. Acad. Sci. USA 9:13,193–97 Appay V, Nixon DF, Donahoe SM, Gillespie GM, Dong T, King A, Ogg GS, Spiegel HM, Conlon C, Spina CA, Havlir DV, Richman DD, Waters A, Easterbrook P, McMichael AJ, Rowland-Jones SL. 2000. HIV-specific CD8+ T cells produce antiviral cytokines but are impaired in cytolytic function. J. Exp. Med. 192:63–75 Champagne P, Ogg GS, King AS, Knabenhans C, Ellefsen K, Nobile M, Appay V, Rizzardi GP, Fleury S, Lipp M, Forster R, Rowland-Jones S, Sekaly RP, McMichael AJ, Pantaleo G. 2001. Skewed maturation of memory HIV-specific CD8 T lymphocytes. Nature 410:106–11 Goulder PJR, Phillips RE, Colbert RA,

53.

54.

55.

56.

57.

58.

59.

McAdam S, Ogg G, Nowak MA, Giangrande P, Luzzi G, Morgan B, Edwards A, McMichael AJ, Rowland-Jones S. 1997. Late escape from an immunodominant cytotoxic T-lymphocyte response associated with progression to AIDS. Nat. Med. 3: 212–17 Zajac AJ, Blattman JN, Murali-Krishna K, Sourdive DJ, Suresh M, Altman JD, Ahmed R. 1998. Viral immune evasion due to persistence of activated T cells without effector function. J. Exp. Med. 188:2205– 13 McMichael AJ, Rowland-Jones SL. 2001. Cellular immune responses to HIV. Nature 410:980–87 Pitcher CJ, Quittner C, Peterson DM, Connors M, Koup RA, Maino VC, Picker LJ. 1999. HIV-1-specific CD4+ T cells are detectable in most individuals with active HIV-1 infection, but decline with prolonged viral suppression. Nat. Med. 5:518– 25 Rosenberg ES, Billingsley JM, Caliendo AM, Boswell SL, Sax PE, Kalams SA, Walker BD. 1997. Vigorous HIV-1–specific CD4+ T cell responses associated with control of viremia. Science 278:1447–50 Rosenberg ES, Altfeld M, Poon SH, Philips MN, Wilkes BM, Eldridge RL, Robbins GK, D’Aquila RT, Goulder PJR, Walker BD. 2000. Immune control of HIV1 after treatment of acute infection. Nature 407:523–26 Mascola JR, Lewis MG, Stiegler G, Harris D, VanCott TC, Hayes D, Louder MK, Brown CR, Sapan CV, Frankel SS, Lu Y, Robb ML, Katinger H, Birx DL. 1999. Protection of macaques against pathogenic simian/human immunodeficiency virus 89.6PD by passive transfer of neutralizing antibodies. J. Virol. 73:4009–18 Mascola JR, Stiegler G, VanCott TC, Katinger H, Carpenter CB, Hanson CE, Beary H, Hayes D, Frankel SS, Birx DL, Lewis MG. 2000. Protection of macaques against vaginal transmission of a pathogenic HIV-1/SIV chimeric virus by passive

4 Feb 2002

11:19

AR

AR152-04.tex

AR152-04.SGM

LaTeX2e(2001/05/10)

P1: GSR

PROSPECTS FOR AIDS VACCINES

Annu. Rev. Immunol. 2002.20:73-99. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

60.

61.

62.

63.

64.

65.

infusion of neutralizing antibodies. Nat. Med. 6:207–10 Hofmann-Lehmann R, Vlasak J, Rasmussen RA, Smith BA, Baba TW, Liska V, Ferrantelli F, Montefiori DC, McClure HM, Anderson DC, Bernacky BJ, Rizvi TA, Schmidt R, Hill LR, Keeling ME, Katinger H, Stiegler G, Cavacini LA, Posner MR Chou T-C, Andersen J, Ruprecht RM. 2001. Postnatal passive immunization of neonatal macaques with a triple combination of human monoclonal antibodies against oral simian-human immunodeficiency virus challenge. J. Virol. 75:70– 80 Letvin NL, Montefiori DC, Yasutomi Y, Perry HC, Davies M-E, Lekutis C, Alroy M, Freed DC, Lord CI, Handt LK, Liu MA, Shiver JW. 1997. Potent, protective anti-HIV immune responses generated by bimodal HIV envelope DNA plus protein vaccination. Proc. Natl. Acad. Sci. USA 94:9378–83 Earl PL, Sugiura W, Montefiori DC, Broder CC, Lee SA, Wild C, Lifson J, Moss B. 2001. Immunogenicity and protective efficacy of oligomeric human immunodeficiency virus type 1 gp140. J. Virol. 75:645–53 Cho MW, Kim YB, Lee MK, Gupta K, Ross W, Plishka R, Buckler-White A, Igarashi T, Theodore T, Byrum R, Kemp C, Montefiori DC, Martin MA. 2001. Polyvalent envelope glycoprotein vaccine elicits a broader neutralizing antibody response but is unable to provide sterilizing protection against heterologous simian/human immunodeficiency virus infection in pigtailed macaques. J. Virol. 75:2224– 34 Feny¨o EM, Albert J, McKeating J. 1996. The role of the humoral immune response in HIV infection. 1996. AIDS 10(suppl. A):S97–S106 Mascola JR, Louwagie J, McCutchan FE, Fischer CL, Hegerich PA, Wagner KF, Fowler AK, McNeil JG, Burke DS. 1994. Two antigenically distinct subtypes of hu-

66.

67.

68.

69.

70.

71.

72.

93

man immunodeficiency virus type 1: viral genotype predicts neutralization serotype. J. Infect. Dis. 169:48–54 Moore JP, Cao Y, Leu J, Qin L, Korber B, Ho DD. 1996. Inter- and intraclade neutralization of human immunodeficiency virus type 1: Genetic clades do not correspond to neutralization serotypes but partially correspond to gp120 antigenic serotypes. J. Virol. 70:427–44 Weber J, Feny¨o EM, Beddows S, Kale˚ 1996. Neutralization ebu P, Bj¨orndal A. serotypes of human immunodeficiency virus type 1 field isolates are not predicted by genetic subtype. J. Virol. 70:7827– 32 Liang X, Munshi S, Shendure J, Mark G III, Davies ME, Freed DC, Montefiori DC, Shiver JW. 1999. Epitope insertion into variable loops of HIV-1 gp120 as a potential means to improve immunogenicity of viral envelope protein. Vaccine 17:2862– 72 Zwick MB, Bonnycastle LLC, Menendez A, Irving MB, Barbas CF III, Parren PWHI, Burton DR, Scott JK. 2001. Identification and characterization of a peptide that specifically binds the human, broadly neutralizing anti-human immunodeficiency virus type 1 antibody b12. J. Virol. 75:6692–99 Back NKT, Smit L, de Jong J-J, Keulen W, Schutten M, Goudsmit J, Tersmette M. 1994. An N-glycan within the human immunodeficiency virus type 1 gp120 V3 loop affects virus neutralization. Virology 199:431–38 Reitter JN, Means RE, Desrosiers RC. 1998. A role for carbohydrates in immune evasion in AIDS. Nat. Med. 4:679– 84 Ye Y, Si ZH, Moore JP, Sodroski J. 2000. Association of structural changes in the V2 and V3 loops of the gp120 envelope glycoprotein with acquisition of neutralization resistance in a simian-human immunodeficiency virus passaged in vivo. J. Virol. 74:11,955–62

4 Feb 2002

11:19

Annu. Rev. Immunol. 2002.20:73-99. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

94

AR

LETVIN

AR152-04.tex

¥

AR152-04.SGM

BAROUCH

¥

LaTeX2e(2001/05/10)

P1: GSR

MONTEFIORI

73. Pilgrim AK, Pantaleo G, Cohen OJ, Fink LM, Zhou JY, Zhou JT, Bolognesi DP, Fauci AS, Montefiori DC. 1997. Neutralizing antibody responses to human immunodeficiency virus type 1 in primary infection and long-term non-progressive infection. J. Infect. Dis. 176:924–32 74. D’Souza MP, Livnat D, Bradac JA, Bridges SH. 1997. Evaluation of monoclonal antibodies to human immunodeficiency virus type 1 primary isolates by neutralization assays: performance criteria for selecting candidate antibodies for clinical trials. J. Infect. Dis. 175:1056–1062 75. Parren PWHI, Burton DR. 2001. The antiviral activity of antibodies in vitro and in vivo. Adv. Immunol. 77:195–262 76. Lasky LA, Nakamura G, Smith DH, Fennie C, Shimasaki C, Patzer E, Berman P, Gregory T, Capon DJ. 1987. Delineation of a region of the human immunodeficiency virus gp120 glycoprotein critical for interaction with the CD4 receptor. Cell 50:975–85 77. Myszka DG, Sweet RW, Hensley P, Brigham-Burke M, Kwong PD, Hendrickson WA, Wyatt R, Sodroski J, Doyle ML. 2000. Energetics of the HIV gp120-CD4 binding reaction. Proc. Natl. Acad. Sci. USA 97:9026–31 78. Rizzuto CD, Wyatt R, Hern´andez-Ramos N, Sun Y, Kwong PD, Hendrickson WA, Sodroski J. 1998. A conserved HIV gp120 glycoprotein structure involved in chemokine receptor binding. Science 280:1949– 53 79. Wu L, Gerard NP, Wyatt R, Choe H, Parolin C, Ruffing N, Borsetti A, Cardoso AA, Desjardin E, Newman W, Gerard C, Sodroski J. 1996. CD4-induced interaction of primary HIV-1 gp120 glycoproteins with the chemokine receptor CCR-5. Nature 384:179–83 80. Salzwedel K, Smith ED, Dey B, Berger EA. 2000. Sequential CD4-coreceptor interactions in human immunodeficiency virus type 1 Env function: soluble CD4 activates Env for coreceptor-dependent fu-

81.

82.

83.

84.

85. 86.

87.

88.

89.

90.

91.

sion and reveals blocking activities of antibodies against cryptic conserved epitopes on gp120. J. Virol. 74:326–33 Berger EA. 1997. HIV entry and tropism: the chemokine receptor connection. AIDS 11(suppl. A):S3–S16 Berger EA, Doms RW, Feny¨o E-M, Korber BTM, Littman DR, Moore JP, Sattentau QJ, Schuitemaker H, Sodrsoki J, Weiss RA. 1998. A new classification for HIV-1. Nature 391:240 Weissenhorn W, Dessen A, Harrison SC, Skehel JJ, Wiley DC. 1997. Atomic structure of the ectodomain from HIV-1 gp41. Nature 387:426–30 Chan DC, Fass D, Berger JM, Kim PS. 1997. Core structure of gp41 from the HIV envelope glycoprotein. Cell 89:263–73 Chan DC, Kim PS. 1998. HIV entry and its inhibition. Cell 93:681–84 McDougal JS, Kennedy MS, Orloff SL, Nicholson JKA, Spira TJ. 1996. Mechanism of human immunodeficiency virus type 1 (HIV-1) neutralization: irreversible inactivation of infectivity by anti-HIV-1 antibody. J. Virol. 70:5236–45 Sullivan N, Sun Y, Sattentau Q, Thali M, Wu D, Denisova G, Gershoni J, Robinson J, Moore J, Sodroski J. 1998. CD4induced conformational changes in the human immunodeficiency virus type 1 gp120 glycoprotein: consequences for virus entry and neutralization. J. Virol. 72:4694–4703 LaCasse RA, Follis KE, Trahey M, Scarborough JD, Littman DR, Nunberg JH. 1999. Fusion-competent vaccines: broad neutralization of primary isolates of HIV. Science 283:357–62 Wyatt R, Kwong PD, Desjardins E, Sweet RW, Robinson J, Hendrickson WA, Sodroski JG. 1998. The antigenic structure of the HIV gp120 envelope glycoprotein. Nature 393:705–11 Wyatt R, Sodroski J. 1998. The HIV-1 envelope glycoproteins: fusogens, antigens, and immunogens. Science 280:1884–88 Kwong PD, Wyatt R, Robinson J, Sweet

4 Feb 2002

11:19

AR

AR152-04.tex

AR152-04.SGM

LaTeX2e(2001/05/10)

P1: GSR

PROSPECTS FOR AIDS VACCINES

Annu. Rev. Immunol. 2002.20:73-99. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

92.

93.

94.

95.

96.

97.

98.

RW, Sodroski J, Hendrickson WA. 1998. Structure of an HIV gp120 envelope glycoprotein in complex with the CD4 receptor and a neutralizing human antibody. Nature 393:648–58 Kwong PD, Wyatt R, Majeed S, Robinson J, Sweet RW, Sodroski J, Hendrickson WA. 2000. Structures of HIV-1 gp120 envelope glycoproteins from laboratoryadapted and primary isolates. Structure 8:1329–39 Muster T, Steindl F, Purtscher M, Trkola A, Klima A, Himmler G, Ruker F, Katinger H. 1993. A conserved neutralization epitope on gp41 of human immunodeficiency virus type 1. J. Virol. 67:6642–47 Mo H, Stamatatos L, Ip JE, Barbas CF, Parren PWHI, Burton DR, Moore JP, Ho DD. 1997. Human immunodeficiency virus type 1 mutants that escape neutralization by human monoclonal antibody IgG1b12. J. Virol. 71:6869–74 Saphire EO, Parren PWHI, Pantophlet R, Zwick MB, Morris GM, Rudd PM Dwek RA, Stanfield RL, Burton DR, Wilson IA. 2001. Crystal structure of neutralizing human IgG against HIV-1: a template for vaccine design. Science. 293:1155–59 Scharf O, Golding H, King LR, Eller N, Frazier D, Golding B, Scott DE. 2001. Immunoglobulin G3 from polyclonal human immunodeficiency virus (HIV) immune globulin is more potent than other subclasses in neutralizing HIV type 1. J. Virol. 75:6558–65 Trkola A, Purtscher M, Muster T, Ballaun C, Buchacher A, Sullivan N, Srinivasa K, Sodroski J, Moore JP, Katinger H. 1996. Human monoclonal antibody 2G12 defines a distinctive neutralization epitope on the gp120 glycoprotein of human immunodeficiency virus type 1. J. Virol. 70:1100–8 Daniel MD, Kirchhoff F, Czajak SC, Sehgal PK, Desrosiers RC. 1992. Protective effects of a live attenuated SIV vaccine with a deletion in the nef gene. Science 258:1938–41

95

99. Whatmore AM, Cook N, Hall GA, Sharpe S, Rud EW, Cranage MP. 1995. Repair and evolution of nef in vivo modulates simian immunodeficiency virus virulence. J. Virol. 69:5117–23 100. Baba TW, Jeong YS, Penninck D, Bronson R, Greene MF, Ruprecht RM. 1995. Pathogenicity of live, attenuated SIV after mucosal infection of neonatal macaques. Science 267:1820–5 101. Baba TW, Liska V, Khimani AH, Ray NB, Dailey PJ, Penninck D, Bronson R, Greene MF, McClure HM, Martin LN, Ruprecht RM. 1999. Live attenuated, multiple deleted simian immunodeficiency virus causes AIDS in infant and adult macaques. Nat. Med. 5:194–03 102. Deacon NJ, Tsykin A, Solomon A, Smith K, Ludford-Menting M, Hooker DJ, McPhee DA, Greenway AL, Ellett A, Chatfield C, Lawson VA, Crowe S, Maerz A, Sonza S, Learmont J, Sullivan JS, Cunningham A, Dwyer D, Dowton D, Mills J. 1995. Genomic structure of an attenuated quasi species of HIV-1 from a blood transfusion donor and recipients. Science 270:988–91 103. Learmont JC, Geczy AF, Mills J, Ashton LJ, Raynes-Greenow CH, Garsia RJ, Dyer WB, McIntyre L, Oelrichs RB, Rhodes RI, Deacon NJ, Sullivan JS. 1999. Immunologic and virologic status after 14 to 18 years of infection with an attenuated strain of HIV-1. New Engl. J. Med. 340:1715–22 104. Greenough TC, Sullivan JL, Desrosiers RC. 1999. Declining CD4 T-cell counts in a person infected with nef-deleted HIV-1. New Engl. J. Med. 340:236–37 105. Murphey-Corb M, Martin LN, DavisonFairburn B, Montelaro RC, Miller M, West M, Ohkawa S, Baskin GB, Zhang JY, Putney SD, Allison AC, Eppstein DA. 1989. A formalin-inactivated whole SIV vaccine confers protection in macaques. Science 246:1293–97 106. Stott EJ. 1991. Anti-cell antibody in macaques. Nature 353:93

4 Feb 2002

11:19

Annu. Rev. Immunol. 2002.20:73-99. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

96

AR

LETVIN

AR152-04.tex

¥

AR152-04.SGM

BAROUCH

¥

LaTeX2e(2001/05/10)

P1: GSR

MONTEFIORI

107. Levine AM, Groshen S, Allen J, Munson KM, Carlo DJ, Daigle AE, Ferre F, Jensen FC, Richieri SP, Trauger RJ, Parker JW, Salk PL, Salk J. 1996. Initial studies on active immunization of HIV-infected subjects using a gp120-depleted HIV-1 Immunogen: long-term follow-up. J. Acquir. Immune Defic. Syndr. 11:351–64 108. Kahn JO, Cherng DW, Mayer K, Murray H, Lagakos S. 2000. Evaluation of HIV1 immunogen, an immunologic modifier, administered to patients infected with HIV having 300 to 549 × 106/L CD4 cell counts: a randomized control trial. JAMA 284:2193–2202 109. Berman PW, Gregory TJ, Riddle L, Nakamura GR, Champe MA, Porter JP, Wurm FM, Hershberg RD, Cobb EK, Eichberg JW. 1990. Protection of chimpanzees from infection by HIV-1 after vaccination with recombinant glycoprotein gp120 but not gp160. Nature 345: 622–25 110. Mascola JR, Snyder SW, Weislow OS, Belay SM, Belshe RB, Schwartz DH, Clements ML, Dolin R, Graham BS, Gorse GJ, Keefer MC, McElrath MJ, Walker MC, Wagner KF, McNeil JG, McCutchan FE, Burke DS. 1996. Immunization with envelope subunit vaccine products elicits neutralizing antibodies against laboratory-adapted but not primary isolates of human immunodeficiency virus type 1. J. Infect. Dis. 173:340–48 111. Conner RI, Korber BTM, Graham BS, Hahn BH, Ho DD, Walker BD, Neumann AU, Vermund SH, Mestecky J, Jackson S, Fenamore E, Cao Y, Gao F, Kalams S, Kunstman KJ, McDonald D, McWilliams N, Trkola A, Moore JP, Wolinksy SM. 1998. Immunological and virological analyses of persons infected by human immunodeficiency virus type 1 while participating in trials of recombinant gp120 subunit vaccines. J. Virol. 72:1552–76 112. Hu S-L, Fultz PN, McClure HM, Eichberg JW, Thomas EK, Zarling J, Singhal

113.

114.

115.

116.

117.

118.

119.

MC, Kosowski SG, Swenson RB, Anderson DC, Todaro G. 1987. Effect of immunization with a vaccinia-HIV env recombinant on HIV infection of chimpanzees. Nature 328:721–23 Shen L, Chen ZW, Miller MD, Stallard V, Mazzara GP, Panicali DL, Letvin NL. 1991. Recombinant virus vaccine— induced SIV-specific CD8+ cytotoxic T lymphocytes. Science 252:440–43 Hu S-L, Abrams K, Barber GN, Moran P, Zarling JM, Langlois AJ, Kuller L, Morton WR, Benveniste RE. 1992. Protection of macaques against SIV infection by subunit vaccines of SIV envelope glycoprotein gp160. Science 255:456–59 Redfield RR, Wright DC, James WD, Jones TS, Brown C, Burke DS. 1987. Disseminated vaccinia in a military recruit with human immunodeficiency virus (HIV) disease. New Engl. J. Med. 316: 673–76 Meyer H, Sutter G, Mayr A. 1991. Mapping of deletions in the genome of the highly attenuated vaccinia virus MVA and their influence on virulence. J. Gen. Virol. 72:1031–38 Hirsch VM, Fuerst TR, Sutter G, Carroll MW, Yang LC, Goldstein S, Piatak M, Elkins W, Alvord WG, Montefiori DC, Moss B, Lifson JD. 1996. Patterns of viral replication correlate with outcome in simian immunodeficiency virus (SIV)infected macaques: effect of prior immunization with a trivalent SIV vaccine in modified vaccinia virus Ankara. J. Virol. 70:3741–52 Ourmanov I, Brown CR, Moss B, Carroll M, Wyatt L Pletneva L, Goldstein S, Venzon D, Hirsch VM. 2000. Comparative efficacy of recombinant modified vaccinia virus Ankara expressing simian immunodeficiency virus (SIV) Gag-Pol and/or Env in macaques challenged with pathogenic SIV. J. Virol. 74:2740–51 Evans TG, Keefer MC, Weinhold KJ, Wolff M, Montefiori D, Gorse GJ, Graham BS, McElrath MJ, Clements-Mann

4 Feb 2002

11:19

AR

AR152-04.tex

AR152-04.SGM

LaTeX2e(2001/05/10)

P1: GSR

PROSPECTS FOR AIDS VACCINES

Annu. Rev. Immunol. 2002.20:73-99. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

120.

121.

122.

123.

124.

125.

126.

ML, Mulligan MJ, Fast P, Walker MC, Excler JL, Duliege AM, Tartaglia J. 1999. A canarypox vaccine expressing multiple human immunodeficiency virus type 1 genes given alone or with rgp120 elicits broad and durable CD8+ cytotoxic T lymphocyte responses in seronegative volunteers. J. Infect. Dis. 180:290–98 Ferrari G, Humphrey W, McElrath MJ, Excler J-L, Duliege A-M, Clements ML, Corey LC, Bolognesi DP, Weinhold KJ. 1997. Clade B-based HIV-1 vaccines elicit cross-clade cytotoxic T lymphocyte reactivities in uninfected volunteers. Proc. Natl. Acad. Sci. USA 94:1396–1401 Sullivan NJ, Sanchez A, Rollin PE, Yang ZY, Nabel GJ. 2000. Development of a preventive vaccine for Ebola virus infection in primates. Nature 30:605–9 Mossman SP, Bex F, Berglund P, Arthos J, O’Neil SP, Riley D, Maul DH, Bruck C, Momin P, Burny A, Fultz PN, Mullins JI, Liljestrom P, Hoover EA. 1996. Protection against lethal simian immunodeficiency virus SIVsmmPBj14 disease by a recombinant Semliki Forest virus gp160 vaccine and by a gp120 subunit vaccine. J. Virol. 70:1953–60 Davis NL, Caley IJ, Brown KW, Betts MR, Irlbeck DM, McGrath KM, Connell MJ, Montefiori DC, Frelinger JA, Swanstrom R, Johnson PR, Johnston RE. 2000. Vaccination of macaques against pathogenic simian immunodeficiency virus with Venezuelan equine encephalitis virus replicon particles. J. Virol. 74:371–78 Hone DM, Wu S, Powell RJ, Pascual DW, Van Cott J, McGhee J, Fouts TR, Tuskan RG, Lewis GK. 1996. Optimization of live oral Salmonella-HIV-1 vaccine vectors for the induction of HIV-specific mucosal and systemic immune responses. J. Biotechnol. 44:203–7 Donnelly JJ, Ulmer JB, Shiver JW, Liu MA. 1997. DNA vaccines. Annu. Rev. Immunol. 15:617–48 Moore JP, Cao Y, Qing L, Sattentau QJ, Pyati J, Koduri R, Robinson J, Bar-

127.

128.

129.

130.

131.

132.

133.

97

bas CF III, Burton DR, Ho DD. 1995. Primary isolates of human immunodeficiency virus type 1 are relatively resistant to neutralization by monoclonal antibodies to gp120, and their neutralization is not predicted by studies with monomeric gp120. J. Virol. 69:101–9 Montefiori DC, Evans TG. 1999. Toward an HIV type 1 vaccine that generates potent, broadly cross-reactive neutralizing antibodies. AIDS Res. Hum. Retroviruses, 15:689–98 Binley JM, Sanders RW, Clas B, Schuelke N, Master A, Guo Y, Kajumo F, Anselma DJ, Maddon PJ, Olson WC, Moore JP. 2000. A recombinant human immunodeficiency virus type 1 envelope glycoprotein complex stabilized by an intramolecular disulfide bond between the gp120 and gp41 subunits is an antigenic mimic of the trimeric virion-associated structure. J. Virol. 74:627–43 Yang X, Florin L, Farzan M, Kolchinsky P, Kwong PD, Sodroski J, Wyatt R. 2000. Modifications that stabilize human immunodeficiency virus envelope glycoprotein trimers in solution. J. Virol. 74: 4746–54 Parren PWHI, Burton DR, Sattentau QJ. 1997. HIV-1 antibody—debris or virion? Nat. Med. 3:366–67 Garrity RR, Rimmelzwaan G, Minassian A, Tsai W-P, Lin G, de Jong J-J, Goudsmit J, Nara PL. 1997. Refocusing neutralizing antibody responses by targeted dampening of an immunodominant epitope. Proc. Natl. Acad. Sci. USA 159:279–89 Rasmussen RA, Montefiori DC, Robinson HL, McClure HM, Ruprecht RM. 2001. Neutralizing antibody induction in a simian-human immunodeficiency virus (SHIV) primate model: lack of original antigenic sin. J. Infect. Dis. In press Barnett SW, Lu S, Sravastava I Cherelis S, Gettie A, Blanchard J, Wang S, Mboudjeka I, Leung L, Lian Y, Fong A, Buckner C, Ly A, Hilt S, Ulmer J,

4 Feb 2002

11:19

98

Annu. Rev. Immunol. 2002.20:73-99. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

134.

135.

136.

137.

138.

139.

AR

LETVIN

AR152-04.tex

¥

AR152-04.SGM

BAROUCH

¥

LaTeX2e(2001/05/10)

P1: GSR

MONTEFIORI

Wild CT, Mascola JR, Stamatatos L. 2001. The ability of an oligomeric human immunodeficiency virus type 1 (HIV1) envelope antigen to elicit neutralizing antibodies against primary HIV-1 isolates is improved following partial deletion of the second hypervariable region. J. Virol. 75:5526–40 Fouts TR, Tuskan R, Godfrey K, Reitz M, Hone D, Lewis GK, DeVico AL. 2000. Expression and characterization of a single-chain polypeptide analogue of the human immunodeficiency virus type 1 gp120-CD4 receptor complex. J. Virol. 74:11,427–436 Scala G, Chen X, Liu W, Telles JN, Cohen OJ, Vaccarezza M, Igarashi T, Fauci AS. 1999. Selection of HIV-specific immunogenic epitopes by screening random peptide libraries with HIV-1-positive sera. J. Immunol. 162:6155–61 Belshe RB, Gorse GJ, Mulligan MJ, Evans TG, Keefer MC, Excler J-L, Duliege A-M, Tartaglia J, McNamara J, Kai-Lin H, Montefiori D, Weinhold K. 1998. Rapid induction of HIV-1 immune responses by canarypox (ALVAC) HIV-1 and gp120 SF2 recombinant vaccines in uninfected Volunteers. AIDS 12:2407–15 Ourmanov I, Bilska M, Hirsch VH, Montefiori DC. 2000. Recombinant modified vaccinia virus Ankara expressing the surface gp120 of simian immunodeficiency virus (SIV) primes for a rapid neutralizing antibody response to SIV infection in macaques. J. Virol. 74:2960–65 Oxenius A, Price DA, Easterbrook PJ, O’Callaghan CA, Kelleher AD, Whelan JA, Sontag G, Sewell AK, Phillips RE. 2000. Early highly active antiretroviral therapy for acute HIV-1 infection preserves immune function of CD8+ and CD4+ T lymphocytes. Proc. Natl. Acad. Sci. USA 97:33–33 Altfeld M, Rosenberg ES, Shankarappa R, Mukherjee JS, Hecht FM, Eldridge RL, Addo MM, Poon SH, Phillips MN, Robbins GK, Sax PE, Boswel S, Kahn

140.

141.

142.

143.

144.

145.

JO, Brander C, Goulder PJ, Levy JA, Mullins JI, Walker BD. 2001. Cellular immune responses and viral diversity in individuals treated during acute and early HIV-1 infection. J. Exp. Med. 193:169–80 Lori F, Lewis MG, Xu J, Varga G, Zinn Jr DE, Crabbs C, Wagner W, Greenhouse J, Silvera P, Yalley-Ogunro J, Tinelli C, Lisziewicz J. 2000. Control of SIV rebound through structured treatment interruptions during early infection. Science 290:1591–93 Ortiz GM, Nixon DF, Trkola A, Binley J, Jin X, Bonhoeffer S, Kuebler PJ, Donahoe SM, Demoitie DA, Kakimoto WM, Ketas T, Clas B, Heymann JJ, Zhang L, Cao Y, Hurley A, Moore JP, Ho DD, Markowitz M. 1999. HIV-1-specific immune responses in subjects who temporarily contain virus replication after discontinuation of highly active antiretroviral therapy. J. Clin. Invest. 104:677–78 Carcelain G, Tubiana R, Samri A, Calvez V, Delaugerre C, Agut H, Katlama C, Autran B. 2001. Transient mobilization of human immunodeficiency virus (HIV)specific CD4 T-helper cells fails to control virus rebounds during intermitent antiretroviral therapy in chronic HIV type 1 infection. J. Virol. 75:234–241 Hel Z, Venzon D, Poudyal M, Tsai WP, Giuliani L, Woodward R, Chougnet C, Shearer G, Altman JD, Watkins D, Bischofberger N, Abimiku A, Markham P, Tartaglia J, Franchini G. 2000. Viremia control following antiretroviral treatment and therapeutic immunization during primary SIV251 infection of macaques. Nat. Med. 6:1140–46 Dhodapkar MV, Steinman RM, Sapp M, Desai H, Fossella C, Krasovsky J, Donahoe SM, Dunbar PR, Cerundolo V, Nixon DF, Bhardwaj N. 1999. Rapid generation of broad T-cell immunity in humans after a single injection of mature dendritic cells. J. Clin. Invest. 104:173–80 Dhodapkar MV, Krasovsky J, Steinman RM, Bhardwaj N. 2000. Mature dendritic

4 Feb 2002

11:19

AR

AR152-04.tex

AR152-04.SGM

LaTeX2e(2001/05/10)

P1: GSR

PROSPECTS FOR AIDS VACCINES

Annu. Rev. Immunol. 2002.20:73-99. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

cells boost functionally superior CD8+ Tcell in humans without foreign helper epitopes. J. Clin. Invest. 105:R9–R14 146. Piatak M Jr, Saag MS, Yang LC, Clark SJ, Kappes JC, Luk KC, Hahn BH, Shaw GM, Lifson JD. 1993. High levels of HIV1 in plasma during all stages of infection determined by competitive PCR. Science 259:1749–54 147. Mellors JW, Rinaldo CR Jr, Gupta P,

99

White RM, Todd JA, Kingsley LA. 1996. Prognosis in HIV-1 infection predicted by the quantity of virus in plasma. Science 272:1167–70 148. Quinn TC, Wawer MJ, Sewankambo N, Serwadda D, Li C, Wabwire-Mangen F, Meehan MO, Lutalo T, Gray RH. 2000. Viral load and heterosexual transmission of human immunodeficiency virus type 1. New Engl. J. Med. 342:921–29

P1: FRK

February 1, 2002

12:47

Annual Reviews

AR152-FM

Annual Review of Immunology Volume 20, 2002

Annu. Rev. Immunol. 2002.20:73-99. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

CONTENTS FRONTISPIECE, Charles A. Janeway, Jr. A TRIP THROUGH MY LIFE WITH AN IMMUNOLOGICAL THEME, Charles A. Janeway, Jr.

xiv 1

THE B7 FAMILY OF LIGANDS AND ITS RECEPTORS: NEW PATHWAYS FOR COSTIMULATION AND INHIBITION OF IMMUNE RESPONSES, Beatriz M. Carreno and Mary Collins

29

MAP KINASES IN THE IMMUNE RESPONSE, Chen Dong, Roger J. Davis, and Richard A. Flavell

PROSPECTS FOR VACCINE PROTECTION AGAINST HIV-1 INFECTION AND AIDS, Norman L. Letvin, Dan H. Barouch, and David C. Montefiori T CELL RESPONSE IN EXPERIMENTAL AUTOIMMUNE ENCEPHALOMYELITIS (EAE): ROLE OF SELF AND CROSS-REACTIVE ANTIGENS IN SHAPING, TUNING, AND REGULATING THE AUTOPATHOGENIC T CELL REPERTOIRE, Vijay K. Kuchroo, Ana C. Anderson, Hanspeter Waldner, Markus Munder, Estelle Bettelli, and Lindsay B. Nicholson

55 73

101

NEUROENDOCRINE REGULATION OF IMMUNITY, Jeanette I. Webster, Leonardo Tonelli, and Esther M. Sternberg

125

MOLECULAR MECHANISM OF CLASS SWITCH RECOMBINATION: LINKAGE WITH SOMATIC HYPERMUTATION, Tasuku Honjo, Kazuo Kinoshita, and Masamichi Muramatsu

165

INNATE IMMUNE RECOGNITION, Charles A. Janeway, Jr. and Ruslan Medzhitov

KIR: DIVERSE, RAPIDLY EVOLVING RECEPTORS OF INNATE AND ADAPTIVE IMMUNITY, Carlos Vilches and Peter Parham ORIGINS AND FUNCTIONS OF B-1 CELLS WITH NOTES ON THE ROLE OF CD5, Robert Berland and Henry H. Wortis E PROTEIN FUNCTION IN LYMPHOCYTE DEVELOPMENT, Melanie W. Quong, William J. Romanow, and Cornelis Murre

197 217 253 301

LYMPHOCYTE-MEDIATED CYTOTOXICITY, John H. Russell and Timothy J. Ley

SIGNAL TRANSDUCTION MEDIATED BY THE T CELL ANTIGEN x

323

P1: FRK

February 1, 2002

12:47

Annual Reviews

AR152-FM

CONTENTS

xi

RECEPTOR: THE ROLE OF ADAPTER PROTEINS, Lawrence E. Samelson INTERACTION OF HEAT SHOCK PROTEINS WITH PEPTIDES AND ANTIGEN PRESENTING CELLS: CHAPERONING OF THE INNATE AND ADAPTIVE IMMUNE RESPONSES, Pramod Srivastava CHROMATIN STRUCTURE AND GENE REGULATION IN THE IMMUNE SYSTEM, Stephen T. Smale and Amanda G. Fisher PRODUCING NATURE’S GENE-CHIPS: THE GENERATION OF PEPTIDES FOR DISPLAY BY MHC CLASS I MOLECULES, Nilabh Shastri, Susan

371

Annu. Rev. Immunol. 2002.20:73-99. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

Schwab, and Thomas Serwold

395 427

463

THE IMMUNOLOGY OF MUCOSAL MODELS OF INFLAMMATION, Warren Strober, Ivan J. Fuss, and Richard S. Blumberg T CELL MEMORY, Jonathan Sprent and Charles D. Surh

GENETIC DISSECTION OF IMMUNITY TO MYCOBACTERIA: THE HUMAN MODEL, Jean-Laurent Casanova and Laurent Abel ANTIGEN PRESENTATION AND T CELL STIMULATION BY DENDRITIC CELLS, Pierre Guermonprez, Jenny Valladeau, Laurence Zitvogel, Clotilde Th´ery, and Sebastian Amigorena

495 551 581

621

NEGATIVE REGULATION OF IMMUNORECEPTOR SIGNALING, Andr´e Veillette, Sylvain Latour, and Dominique Davidson

669

CPG MOTIFS IN BACTERIAL DNA AND THEIR IMMUNE EFFECTS, Arthur M. Krieg

PROTEIN KINASE Cθ

709 IN

T CELL ACTIVATION, Noah Isakov and Amnon

Altman

RANK-L AND RANK: T CELLS, BONE LOSS, AND MAMMALIAN EVOLUTION, Lars E. Theill, William J. Boyle, and Josef M. Penninger PHAGOCYTOSIS OF MICROBES: COMPLEXITY IN ACTION, David M. Underhill and Adrian Ozinsky

761 795 825

STRUCTURE AND FUNCTION OF NATURAL KILLER CELL RECEPTORS: MULTIPLE MOLECULAR SOLUTIONS TO SELF, NONSELF DISCRIMINATION, Kannan Natarajan, Nazzareno Dimasi, Jian Wang, Roy A. Mariuzza, and David H. Margulies

853

INDEXES Subject Index Cumulative Index of Contributing Authors, Volumes 1–20 Cumulative Index of Chapter Titles, Volumes 1–20

ERRATA An online log of corrections to Annual Review of Immunology chapters (if any, 1997 to the present) may be found at http://immunol.annualreviews.org/

887 915 925

8 Feb 2002

9:37

AR

AR152-05.tex

AR152-05.SGM

LaTeX2e(2001/05/10) P1: GSR 10.1146/annurev.immunol.20.081701.141316

Annu. Rev. Immunol. 2002. 20:101–23 DOI: 10.1146/annurev.immunol.20.081701.141316 c 2002 by Annual Reviews. All rights reserved Copyright °

Annu. Rev. Immunol. 2002.20:101-123. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

T CELL RESPONSE IN EXPERIMENTAL AUTOIMMUNE ENCEPHALOMYELITIS (EAE): Role of Self and Cross-Reactive Antigens in Shaping, Tuning, and Regulating the Autopathogenic T Cell Repertoire Vijay K. Kuchroo,1 Ana C. Anderson,1 Hanspeter Waldner,1 Markus Munder,2 Estelle Bettelli,1 and Lindsay B. Nicholson1 1

Center for Neurologic Diseases, Brigham and Women’s Hospital and Harvard Medical School, 77 Avenue Louis Pasteur, Boston, Massachusetts 02115; e-mail: [email protected]; [email protected]; [email protected]; [email protected]; [email protected] 2 Medizinische Klinik und Poliklinik V, Ruprecht-Karls-Universit¨at, Heidelberg, Hospitalstrasse 3, 69115 Heidelberg, Germany; e-mail: [email protected]

Key Words myelin antigens (MBP, PLP), thymic selection, selective tolerance, susceptibility/resistance, pathogenic/protective repertoires ■ Abstract T cells that can respond to self-antigens are present in the peripheral immune repertoire of all healthy individuals. Recently we have found that unmanipulated SJL mice that are highly susceptible to EAE also maintain a very high frequency of T cells responding to an encephalitogenic epitope of a myelin antigen proteolipid protein (PLP) 139-151 in the peripheral repertoire. This is not due to lack of expression of myelin antigens in the thymus resulting in escape of PLP 139-151 reactive cells from central tolerance, but is due to expression of a splice variant of PLP named DM20, which lacks the residues 116-150. In spite of this high frequency, the PLP 139-151 reactive cells remain undifferentiated in the periphery and do not induce spontaneous EAE. In contrast, SJL TCR transgenic mice expressing a receptor derived from a pathogenic T cell clone do develop spontaneous disease. This may be because in normal mice, autoreactive cells are kept in check by an alternate PLP 139-151 reactive nonpathogenic repertoire, which maintains a balance that keeps them healthy. If this is the case, selective activation of one repertoire or the other may alter susceptibility to autoimmune disease. Since T cells are generally cross-reactive, besides responding to nonself-antigens, they also maintain significant responses to self-antigens. Based on the PLP 139-151 system, we propose a model in which activation with foreign antigens can 0732-0582/02/0407-0101$14.00

101

4 Feb 2002

11:21

102

AR

AR152-05.tex

AR152-05.SGM

LaTeX2e(2001/05/10)

P1: GSR

KUCHROO ET AL.

result in the generation of pathogenic memory T cells that mediate autoimmunity. We also outline circumstances under which activation of self-reactive T cells with foreign antigens can generate selective tolerance and thus generate protective/regulatory memory against self while still maintaining significant responses against foreign antigens. This provides a mechanism by which the fidelity and specificity of the immune system against foreign antigens is improved without increasing the potential for developing an autoimmune disease.

Annu. Rev. Immunol. 2002.20:101-123. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

INTRODUCTION Animals immunized with components of central nervous system (CNS) myelin develop an autoimmune demyelinating disease of the CNS called experimental autoimmune encephalomyelitis (EAE) (1–3). This disease is the result of a CD4+ T cell–mediated immune response directed at specific proteins within the CNS, and it serves as a model of the human disease multiple sclerosis (MS). In the EAEsusceptible mouse strain SJL (H-2s), the immune response following immunization with whole spinal cord homogenate is focused on a single peptide epitope, myelin proteolipid protein (PLP) peptide 139-151 (4). This strikingly dominant immune response is mediated by T cells with a broad repertoire of different T cell receptors (TCRs) (5). The investigation of this diverse, dominant response in our laboratory, and in others, has led to several novel insights into the role of self- and cross-reactive antigens in shaping, tuning, and regulating the autoreactive T cell repertoire, which are the focus of this review. Appreciating the pivotal role of autoreactive CD4+ T cells in the development of organ-specific autoimmune disease marked a significant advance in our understanding of the pathogenesis of autoimmunity. Most autoreactive T cells are deleted during thymic development, and this seems to be true even for T cells specific for tissue antigens (like myelin antigens), because many of these antigens are now known to be expressed in the thymus during ontogeny. However, thymic deletion of autoreactive T cells cannot be complete, and the reasons for this are discussed below. Autoreactive T cells that escape thymic deletion constitute the peripheral T cell repertoire and are kept in check in the circulation by the mechanisms of peripheral tolerance. Autoreactive T cells in the periphery cannot become activated until they encounter antigen in the context of relevant major histocompatibility complex (MHC) molecules, together with appropriate costimulatory signals. In CNS inflammatory diseases like EAE and MS, the T cells then migrate across the blood brain barrier to cause disease (6). In the CNS, autoreactive T cells have to be activated by local antigen presenting cells (APCs), also expressing appropriate costimulatory molecules, to initiate inflammation and tissue injury (7). It is generally thought that EAE, and by inference MS, is a DTH-type reaction driven by T cells that have differentiated to a Th1 phenotype (8). Although myelin antigen-specific Thl cells are necessary to initiate the disease, most of the cells seen in EAE lesions are recruited nonspecifically. These infiltrating cells consist mainly of T cells and macrophages and, to a lesser extent, B cells. In some cases, polymorphs are also detected in acute EAE lesions

4 Feb 2002

11:21

AR

AR152-05.tex

AR152-05.SGM

LaTeX2e(2001/05/10)

P1: GSR

Annu. Rev. Immunol. 2002.20:101-123. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

T CELL RESPONSE IN EAE

103

(9). Nonspecifically recruited cells are thought to play a major role in the tissue damage. Activated macrophages strip myelin from axons and secrete numerous cytokines including IL-l and TNF-α, which can perpetuate nonspecific inflammatory reactions and contribute to tissue damage. Therefore, following escape from the thymus, an autoreactive T cell has to undergo several discrete steps to mediate an autoimmune disease: The T cells have to be activated in the peripheral immune compartment, differentiate to attain a pathogenic effector phenotype, express appropriate adhesion molecules to traffic to the target organ, and get reactivated to recruit other cells to mediate tissue injury and develop autoimmune disease. How thymic expression of tissue myelin antigens, and how reactivity with self- and cross-reactive antigens in the peripheral immune compartment, might shape, tune, and regulate the autoreactive T cell repertoire are discussed below. There is growing evidence that T cells that can respond to self-antigens are present in the peripheral T lymphocyte repertoire of all humans. T cells reactive with myelin basic protein (MBP) (10), myelin proteolipid protein (PLP) (11), and other self-antigens can readily be expanded and cloned from the peripheral blood of healthy individuals. To dissect the selection of the autoreactive repertoire and its regulation/activation in the periphery in a context relevant to human disease, it is useful to have a model in which naive T cells cross-reactive with a known autoantigen are generated at high frequencies. Mice expressing MHC H-2s provide such a model. Recently, we have shown that in the unmanipulated SJL (H-2s) strain, which is highly susceptible to EAE, the frequency of T cells reactive to the encephalitogenic epitope of PLP, PLP 139-151, is as high as 1/20,000 in na¨ıve animals (12). The presence of this large, diverse, endogenous self-reactive repertoire provides an explanation for the immunodominance of this epitope. We have also developed a TCR transgenic model in which the majority of T cells recognize PLP 139-151 (13). The study of both na¨ıve and TCR transgenic H-2s mice has allowed us to analyze how a high frequency of self-reactive cells escapes from the thymus, what the contribution of antigen (self- or cross-reactive) and MHC haplotype are to the selection process, and what the factors are that control the activation of autopathogenic naive and memory cells in the periphery. Since normal SJL mice remain healthy unless challenged, and transgenic mice develop spontaneous disease, this facilitates the investigation of the mechanisms that tip the balance between tolerance and autoimmunity in individuals with a T cell repertoire biased toward self-reactivity. The first step in this process is to understand how self-reactive cells escape thymic deletion and/or may even be selected in high numbers in the thymus of susceptible individuals.

THYMIC SELECTION OF THE SELF-REACTIVE REPERTOIRE T cells recognize peptide fragments derived from protein antigens presented in the context of major histocompatibility complex molecules (14, 15). Because the MHC molecules in the thymus present predominantly self-peptides, positive selection gives rise to T cells that are inherently self-reactive. Thus, a process of

4 Feb 2002

11:21

Annu. Rev. Immunol. 2002.20:101-123. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

104

AR

AR152-05.tex

AR152-05.SGM

LaTeX2e(2001/05/10)

P1: GSR

KUCHROO ET AL.

negative selection is required to eliminate highly self-reactive T cells from the pool of positively selected cells, leaving a population of cells in which the ability to discriminate self from foreign antigen is optimal. The fact that the MHC molecules in the thymus present peptides derived from self-proteins implies that for every T cell that recognizes a foreign antigen there is at least one selecting cross-reactive self-peptide. Clearly this poses a problem for understanding self-/nonself-discrimination as it allows for the possibility that cells that recognize foreign antigen may in some cases trigger an autoimmune response. This problem may be compounded if a particular self-antigen is not expressed in the thymus and cannot mediate negative selection or if the MHC/peptide combination is inefficient at mediating negative selection. Two major hypotheses have been advanced to explain the presence of selfreactive cells in the peripheral T cell repertoire, and both postulate a failure in thymic negative selection. The first proposes that many tissue-specific antigens are not expressed in the thymus, thereby precluding any central tolerance to these antigens. The second proposes that inefficient antigen presentation by certain disease-associated MHC alleles results in a peripheral repertoire biased toward autoreactivity. Both of these hypotheses are discussed below.

Impact of Self-Antigen on Selection of the Self-Reactive Repertoire The hypothesis that organ-specific autoimmunity is the result of a failure in central tolerance due to the sequestration of tissue-specific antigens behind anatomical barriers has a long history. It was suggested that a breakdown in these barriers that allows the influx of T cells and the subsequent recognition of these nonthymic antigens as foreign result in the induction of an autoimmune disease. This has been a particularly attractive idea for CNS autoimmunity, where the anatomy of the blood-brain barrier and the lack of lymphatic drainage from the CNS (16, 17) have been thought to preclude the development of central tolerance to myelin antigens such as MBP and PLP. However, recent data indicating that both MBP and PLP are expressed in the thymus require a radical re-examination of this hypothesis (18–22). There also appears to be a case for the expression in the thymus of other “tissuespecific” self-antigens implicated in other organ-specific autoimmune diseases such as diabetes (glutamic acid decarboxylase 65 and insulin), uveitis (retinal S-antigen and interphotoreceptor retinoid binding protein), and thyroiditis (thyroid peroxidase and thyroglobulin) (23–26). Furthermore, as is the case in experimental autoimmune uveitis, variation in the level of “tissue-specific” antigen expression in the thymus in different strains correlates with the degree of susceptibility to the induction of this autoimmune disease (27). The apparent promiscuous expression of “tissue-specific” antigens makes less tenable the hypothesis that a failure of negative selection due to a lack of antigen expression is responsible for the escape of self-reactive cells from the thymus, and instead it suggests that a more complex mechanism is involved.

4 Feb 2002

11:21

AR

AR152-05.tex

AR152-05.SGM

LaTeX2e(2001/05/10)

P1: GSR

Annu. Rev. Immunol. 2002.20:101-123. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

T CELL RESPONSE IN EAE

105

If sequestration alone cannot explain the escape of self-reactive cells from the thymus, several other mechanisms may: (a) “tissue-specific” self-antigens do mediate deletion of high-affinity cells, and the cells that are seeded to the periphery have low affinity for self-antigen compared with their affinity for foreign antigen; (b) self-reactive cells escape because the form of self-antigen expressed in the thymus differs from that expressed in peripheral tissue; (c) “tissue-specific” selfantigens do not mediate negative selection due to inefficient antigen presentation by the MHC alleles present in the individual. Data supporting the first mechanism come from studies of the T cell repertoire to MBP that is generated in wild-type and MBP-deficient shiverer mice. In one study, MBP-deficient C3H (H-2k) mice immunized with whole MBP exhibited a strong immune response directed at MBP 79-87 (28). In contrast, wild-type C3H mice responded poorly to immunization with whole MBP and generated a T cell response after immunization with MBP 79-87 that is at least 3 logs lower in avidity when compared to the response elicited by MBP 79-87 in MBP-deficient C3H mice. Similarly, another study found that while wild-type Balb/c mice did not respond to immunization with whole MBP or its peptides, MBP-deficient Balb/c mice immunized with whole MBP exhibited a strong immune response directed at two epitopes of MBP, MBP 59-76 and 89-101 (29). In addition, the MBP-reactive cells that are elicited in MBP-deficient Balb/c mice represent disease-inducing cells, as they are able to transfer disease to wild-type Balb/c mice. Although not formally shown in either study, the data suggest that the expression of MBP in wild-type mice results in the deletion of cells that recognize MBP with high avidity. Examples of the second mechanism come from studies of the T cell repertoire to another myelin antigen, PLP. The predominant form of PLP expressed in the thymus is DM20, a splice variant that lacks residues 116-150 of full-length PLP (12, 30). Thus, only those cells that react to epitopes present in DM20 would be expected to undergo central tolerance. Analysis of the T cell repertoire to PLP in wild-type and PLP-deficient C57BL/6 mice shows that cells reactive to epitopes contained within DM20 undergo tolerance in wild-type mice (30). Furthermore, PLP-deficient C57BL/6 mice that are thymectomized, grafted with a wild-type C57BL/6 thymus, and subsequently lethally irradiated and reconstituted with PLPdeficient bone marrow are tolerant to PLP, indicating that the expression of PLP by radio-resistant thymic epithelial cells is sufficient to induce tolerance. Our study of the T cell repertoire to PLP in H-2s mice adds a layer of complexity to these observations. In SJL mice, which are highly susceptible to EAE induced by PLP 139-151, the frequency of T cells reactive to this epitope, which is not contained within DM20, is as high as 1/20,000 in na¨ıve animals (12). Indeed, PLP 139-151reactive cells are already present at a high frequency in the thymocyte population from neonatal mice, and reexpression of PLP 139-151 in the thymuses of embryonic SJL mice results in a significant decrease in this population. These data are in agreement with the conclusions of the first study. However, other data suggest that this lack of PLP 139-151 expression in the thymus and escape from thymic deletion of PLP-reactive T cells may not solely explain the high frequency of PLP

4 Feb 2002

11:21

Annu. Rev. Immunol. 2002.20:101-123. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

106

AR

AR152-05.tex

AR152-05.SGM

LaTeX2e(2001/05/10)

P1: GSR

KUCHROO ET AL.

139-151-reactive T cells in the peripheral repertoire. First, a lack of negative selection alone cannot explain the generation of a frequency of PLP 139-151-reactive cells that is high enough to be detected without immunization. Second, PLP 139151-reactive cells are present at similar frequencies in PLP knockout mice and in mice bred in a germ-free environment, suggesting that these cells are positively selected at a high frequency. Third, PLP 139-151-reactive cells in the periphery are enriched in the CD44 high population, indicating that they have undergone expansion (12). As a whole the above data suggest that in H-2s mice, in addition to the lack of negative selection, cross-reactive self-antigen(s) positively select a high frequency of PLP 139-151-reactive cells that is further expanded in the periphery. An alternative hypothesis is that, as in other systems, the selected repertoire may be rather “lumpy” (31), and for undefined structural reasons, TCRs that react to PLP 139-151 are generated at a high frequency in H-2s mice; this frequency is further augmented by homeostatic mechanisms in the periphery. The third mechanism to explain the presence of autoreactive T cells, inefficient antigen presentation, has been invoked to explain the escape from tolerance of self-reactive cells recognizing the immunodominant encephalitogenic epitope of MBP in H-2u strains, MBP 1-11. This peptide is known to bind weakly to the MHC (IC 50 of 7.4 µM) (32) and to form unstable peptide/MHC complexes (33), which cannot mediate efficient negative selection (34). In addition in MBP-deficient H-2u mice (shiverer mice), the immunodominant epitope of MBP is MBP 121150, which forms highly stable complexes with I-Au (33), suggesting that the expression of MBP in the thymus does result in tolerance to epitopes that form stable peptide/MHC complexes; meanwhile cells reactive to epitopes that form unstable peptide/MHC complexes escape. In the thymus, MBP 1-11 is expressed as part of the embryonic golli-MBP (34a), whereas MBP 121-150 is not (18, 21). To address the question of how cells reactive to MBP 121-150 undergo tolerance in normal mice, Goverman and colleagues generated TCR transgenic mice specific for MBP 121-150. They demonstrated that these cells are in fact deleted in the thymus of normal mice as a result of the presentation of MBP 121-150 by bone marrow–derived APCs that acquire MBP 121-150 exogenously (35). Studies in the NOD mouse model of autoimmune diabetes provide another example of how inefficient antigen presentation may be responsible for the escape of self-reactive cells from the thymus. In the NOD mouse, homozygosity for I-Ag7 is strongly associated with susceptibility to diabetes. In fact, if another I-A allele is introduced onto the NOD background, the incidence of diabetes is significantly reduced (36). NOD mice have also been described as being unable to maintain self-tolerance after immunization with self-peptide in adjuvant, a phenomenon referred to as “autoproliferation” (37). This results in proliferative responses to endogenously processed and presented self-peptides and has been demonstrated to be a property of I-Ag7 (38). The reduction of I-Ag7 expression as a result of the introduction of non-I-Ag7 alleles onto the NOD background leads to a decrease in “autoproliferation” (39). This provides one explanation for the requirement of I-Ag7 homozygosity for the development of autoimmune diabetes in NOD mice.

4 Feb 2002

11:21

AR

AR152-05.tex

AR152-05.SGM

LaTeX2e(2001/05/10)

P1: GSR

T CELL RESPONSE IN EAE

107

In addition, I-Ag7 has been demonstrated to be structurally unstable and to bind some disease-related self-peptides poorly (40–43). Collectively, these observations support a model in which inefficient peptide binding by I-Ag7 results in poor thymic deletion of self-reactive cells in the thymus, which then leads to a failure in maintaining self-tolerance in the periphery (44). The difference between the results in this model and those described above is that poor antigen presentation is not a global property of I-Au.

Annu. Rev. Immunol. 2002.20:101-123. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

Association of Autoimmune Disease with the MHC The strong genetic association between certain MHC alleles and specific autoimmune diseases is well known, and this is perhaps not surprising given the role of the MHC in shaping the T cell repertoire. Since this fact was first discovered, our knowledge of the biochemistry of MHC molecules and the nature of their interaction with the TCR has advanced significantly, but the basis of this association remains poorly understood. One hypothesis, that only disease-associated MHC alleles bind and present self-antigens efficiently, no longer seems tenable given the data demonstrating that all MHC molecules are usually filled with selfantigens. More recent hypotheses suggest that disease-associated alleles shape the self-reactive repertoire in a way that favors the development of autoimmunity. One proposal is that disease-associated MHC alleles present self-antigen poorly and therefore cannot purge self-reactive T cells from the repertoire efficiently. As discussed above, the studies of the association of I-Ag7 with autoimmunity in the NOD mouse model (44) and of MBP 1-11 induced autoimmunity in H-2u strains (33) provide examples of how defects in antigen presentation can result in a repertoire biased toward autoimmunity. Another hypothesis arising from our study of PLP-induced EAE in SJL mice is that certain MHC molecules select T cells that recognize specific autoantigens at very high frequencies (12), increasing the probability that bystander or cross-reactive activation of T cells in the periphery will trigger autoimmunity directed at these particular epitopes. These animals have a peripheral repertoire poised for the development of autoimmune disease. What then needs to be understood are the mechanisms that keep the autoreactive cells in check in susceptible individuals, and the steps that lead to the activation and expansion of pathogenic T lymphocytes in the periphery of animals predisposed to develop autoimmune disease.

PERIPHERAL ACTIVATION OF AUTOPATHOGENIC T CELLS Once seeded to the peripheral immune compartment, autoreactive T cells are normally self-tolerant and do not initiate an autoimmune disease. How these cells remain self-tolerant has been studied at some length by us in the PLP 139-151 system, and at least two factors may play a role: the PLP 139-151-reactive T cells

4 Feb 2002

11:21

Annu. Rev. Immunol. 2002.20:101-123. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

108

AR

AR152-05.tex

AR152-05.SGM

LaTeX2e(2001/05/10)

P1: GSR

KUCHROO ET AL.

in the naive repertoire have not proliferated to an effector phenotype, even though they express high levels of CD44, and there are nonpathogenic PLP 139-151reactive cells in the repertoire that may keep the pathogenic T cells under check (discussed below). Furthermore different H-2s mouse strains differ in their susceptibility to the induction of EAE. In SJL mice, activation of PLP 139-151-specific cells (by PLP 139-151 peptide immunization) is sufficient to induce EAE, and the breakdown of self-tolerance, and thus development of autoimmune disease, only requires activation of the autoreactive T cells in the peripheral immune compartment. On the other hand H-2s-congenic B10.S mice, which also express high numbers of PLP 139-151-reactive T cells (albeit lower than SJL) in the peripheral naive repertoire, are relatively resistant to the development of EAE, even after immunization with PLP 139-151; thus, in the B10.S strain, activation with antigen in CFA is often not sufficient to overcome self-tolerance and mediate EAE (45). This difference in EAE susceptibility between H-2s-congenic SJL and B10.S mice is due to many genetic and environmental factors ultimately affecting expansion and differentiation of self-reactive T cells, an area intensively studied in recent years.

T Cell Differentiation and Susceptibility to EAE The differentiation of naive CD4+ T helper cells into Th1 or Th2 subsets has profound effects on the outcome of autoimmune and infectious disease (46). However, the factors that determine whether an immune response will be dominated in vivo by Th1 or Th2 cells are not well understood. Naive CD4+ T lymphocytes triggered by antigen differentiate into at least two subpopulations, each producing its own set of cytokines and mediating distinct effector functions (47, 48). Type 1 helper T cells (Th1 cells) produce interleukin 2 (IL-2), tumor necrosis factor β (TNF-β), and interferon-gamma (IFN-γ ), activate macrophages, and induce delayed type hypersensitivity (DTH) responses. Type 2 (Th2) cells produce IL-4, IL-5, and IL-10, stimulate the production of mast cells, eosinophils, and IgE antibodies, and regulate cell-mediated immunity (49). IL-4 and IFN-γ show reciprocal inhibition, and IL-10 inhibits the production of IFN-γ and other Th1 cytokines by interfering with antigen presentation by macrophages (50). These two cell populations cross-regulate one another because their respective cytokines act antagonistically (51). A variety of experimental models have shown that the production of IL-12 by macrophages and IFN-γ by natural killer cells promotes the differentiation of naive T cells into Th1 cells and inhibits their differentiation into Th2 cells. Conversely, IL-4 is necessary for Th2 differentiation and inhibits the development of IFN-γ -secreting cells (52). Cytokines play a pivotal role in the initiation, propagation, and regulation of tissue-specific autoimmune injury. Cellular and cytokine changes in the CNS have been described in several studies of MBP-induced EAE. Inflammation appears to precede the clinical signs of EAE; CD4+ cells predominate initially, but peripherally derived macrophages outnumber CD4+ cells as the disease progresses. The autoreactive T cells that induce EAE generally display a Th1 phenotype (53).

4 Feb 2002

11:21

AR

AR152-05.tex

AR152-05.SGM

LaTeX2e(2001/05/10)

P1: GSR

Annu. Rev. Immunol. 2002.20:101-123. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

T CELL RESPONSE IN EAE

109

We (5) and others (6, 54, 55) have generated T cell clones specific for different encephalitogenic epitopes of MBP or PLP. The adoptive transfer of these myelin antigen-reactive Th1 cells can induce EAE. In addition, during progression of EAE, Th1 cytokines are present in the inflammatory EAE lesions in the CNS, whereas Th2 cytokines are absent, strongly suggesting that Th1 cytokines play a role in the pathogenesis of the disease (56, 57). PLP-reactive T cells generated in the presence of IL-12, a cytokine known to enhance the generation of Th1 cells, transferred potent EAE; the PLP-reactive T cells grown in the presence of anti-IL-12 were not encephalitogenic (58). These in vivo studies support the conclusion that Thl cytokines have a role in the pathogenesis of EAE. On the other hand, regulatory T cells that suppress the development of EAE produce cytokines that correspond to the Th2 and Th3 (cells that produce predominantly TGF-β) (59, 60) profile; recovery from EAE in mice and rats is associated with an increase in the presence of Th2 and Th3 cells and cytokines in the CNS (56, 61). These findings, along with the observation that Th2 cytokines can inhibit the actions of inflammatory Th1 cytokines, suggest that the induction and activation of Th2 cells may prevent EAE and other autoimmune diseases mediated by Th1 cells. Indeed, Racke et al. have shown that IL-4-induced immune deviation can be used as a therapy in EAE (62). We have addressed the role of Th2 cells in EAE by generating PLP 139-151reactive Th2 clones by two approaches, immunization in the presence of anti-B7-1 antibody (63) and immunization with altered peptide ligands (APLs) (64). The Th2 clones produced by administering anti-B7-1 antibody together with PLP 139151 immunization in vivo secreted both IL-4 and IL-10; they inhibited EAE if given at the time of immunization and reversed disease if given at first signs of EAE. T cell lines generated with APLs in which primary and/or secondary T cell contact residues were replaced, called Q144 or L144/R147, also induced a Th2/Th0 phenotype and conferred protection in SJL mice immunized for EAE induction with PLP 139-151. Thus, Th2 cells can function as regulators of EAE; however, it was not clear which Th2 cytokines were most important in regulating EAE, or whether these cytokines have equal or overlapping functions in inhibiting EAE. To address this issue in additional studies, we (65) and others (66) have proposed a key role for an IL-12/IL-10 immunoregulatory circuit in the induction of EAE. We demonstrated that IL-10-deficient (IL-10−/−) mice are highly susceptible to the development of EAE and show enhanced Th1 responses and IFN-γ production, whereas IL-10 transgenic mice are highly resistant to EAE induction. This was later confirmed by others (67). In contrast the IL-4−/− mice showed only marginally increased EAE severity, which thus suggests a more important role of IL-10 over IL-4 in the induction of EAE. These data could be explained by the IL-12/IL-10 immunoregulatory circuit in which loss of the inhibitory effects of IL-10 in the deficient mice leads to an increase in IL-12 production and enhanced IFN-γ production from the responding T cells. Studies with EAE-resistant H-2s-congenic B10.S mice also suggest an important role of T cell expansion and differentiation in genetic resistance to EAE.

8 Feb 2002

10:16

Annu. Rev. Immunol. 2002.20:101-123. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

110

AR

AR152-05.tex

AR152-05.SGM

LaTeX2e(2001/05/10)

P1: GSR

KUCHROO ET AL.

Shevach et al. found that resistance to EAE in B10.S mice correlated with an antigen-specific defect in the generation of Th1 cells that produce IFN-γ (66). The T cells from SJL or B10.S mice immunized with MBP proliferated equally well to MBP, but the B10.S T cells did not produce IFN-γ . The defect could be overcome by exposing the MBP-reactive T cells from B10.S mice to exogenous IL-12, giving the cells the ability to transfer EAE into naive recipients (68). In addition, exposure of developing T cell lines from B10.S mice to infectious agents that induce IL-12 also resulted in the generation of an encephalitogenic T cell phenotype (69). Both these pieces of data support the key role of IL-12 in the induction of an encephalitogenic T cell response and a possible MBP-specific defect in the production of IL-12 in the resistant B10.S mice. Our data with PLP 139-151-induced EAE in the SJL and B10.S mouse strains suggest that immunization with PLP 139151 does not induce comparable expansion of PLP 139-151-specific T cells in the B10.S mice as compared to susceptible SJL mice. This might relate to the allelic variation of IL-2 that we have observed between SJL and B10.S mice (70). Furthermore, PLP 139-151-immunized B10.S mice do not produce as much IFN-γ , but they produce anti-inflammatory cytokines such as IL-10 and TGF-β (71). However, although the vast majority of data in EAE suggests that Th1 cells and cytokines are pathogenic and Th2 cells may be protective, there are circumstances under which Th2 cells with specificity for myelin antigens have been shown not to be protective or to be pathogenic, particularly in an immunodeficient host (72, 73).

Two Different Autoreactive T Cell Repertoires with Two Different Functions Because in several systems antigen-specific Th2 T cells are not sufficient to induce disease protection, and to explain why some T cells that react with PLP 139-151 are pathogenic and others are nonpathogenic, we have undertaken an exhaustive analysis of the T cell response to this peptide. When a panel of pathogenic and nonpathogenic T cell clones was analyzed, we found that they expressed a wide range of different TCRs (5) (and unpublished data). Despite this apparent lack of restriction at the TCR level, pathogenic cells all recognized the MHC-peptide similarly, in that they all used a tryptophan (W) at position 144 as the primary TCR contact and L145 and P148 as the MHC contact residues (74). Any substitution at position 144, even with a conservative amino acid did not activate pathogenic T cell clones, supporting the role of W144 in activation of pathogenic T cells. This result has been confirmed independently in three other laboratories (75–77). In contrast, nonpathogenic cells that were produced either by immunization in the presence of anti-B7-1 antibody or by immunization with APLs were drawn from a different population of T cells than the pathogenic cells. They were more cross-reactive with peptides altered at position 144 than were the pathogenic EAE-inducing clones (78). These Th2 clones could tolerate an alanine substitution for tryptophan at position 144, and their recognition of PLP 139-151 depended on leucine (L) at position 141 and glycine (G) at position 142 (Figure 1A).

8 Feb 2002

10:18

AR

AR152-05.tex

AR152-05.SGM

LaTeX2e(2001/05/10)

P1: GSR

Annu. Rev. Immunol. 2002.20:101-123. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

T CELL RESPONSE IN EAE

111

Therefore, expansion of one repertoire that is dependent on W144 in the center of the peptide and that differentiates predominantly along the Th1 pathway induces EAE. The other repertoire is more dependent on the N-terminal residues L141/G142, and this repertoire preferentially differentiates along the Th2 pathway and regulates EAE. These two PLP 139-151-specific repertoires may play a crucial role in balancing the immune responses to PLP 139-151 and in preventing autoimmunity in the normal SJL mouse. Preferential expansion of one or the other repertoire may result in autoimmunity or inhibition/prevention of autoimmune disease (Figure 1B). Since there are at least two functionally different types of T cells specific for PLP 139-151, we cloned the genes from the TCRs of T cell clones that represent each of the populations and used them to generate TCR transgenic mice. Two TCR transgenic lines (4E3 and 5B6), generated with the TCRs from pathogenic clones that depend on tryptophan in the center of the PLP 139-151 peptide for activation, developed fulminant spontaneous EAE (13). In SJL mice the PLP-TCR transgenic T cells developed a Th1 phenotype and induced EAE spontaneously at a very high frequency (45%–60%). In contrast to the PLP-TCR transgenic mice, normal SJL mice, which have a very high frequency of PLP 139-151-reactive T cells in the peripheral repertoire, do not develop spontaneous EAE. Therefore the overexpression of pathogenic cells in the PLP-TCR transgenic mice must overcome normal regulatory mechanisms, which leads to the development of spontaneous disease, perhaps in part because of the lack or reduction of the alternate (regulatory) T cell repertoire. We have established the PLP 139-151-specific TCR transgenic mice (5B6 line) on SJL and B10.S backgrounds. It is interesting that the expression of pathogenic TCRs on the B10.S genetic background induces spontaneous EAE at a much lower frequency than on the SJL background. The lack of spontaneous disease in the B10.S mice is not due to the deletion of transgenic T cells on this background, but appears to be due to functional differences between the two strains. This strongly suggests that at least some of the genetic factors that confer resistance to B10.S mice are independent of the presence of pathogenic T cells (H.Waldner, manuscript in preparation). We have recently generated another TCR transgenic mouse utilizing the TCR from a T cell clone Q1.1B6, which represents the alternate TCR repertoire reactive to PLP 139-151 (79). This T cell can be activated with the PLP 139-151 peptide but is more broadly cross-reactive and does not require tryptophan at position 144 for activation (Figure 1A). The 1B6 TCR transgenic mice do not develop spontaneous EAE when housed in the same room as the 5B6 TCR transgenic mice that develop EAE with a high incidence. Furthermore, the 1B6 transgenic mice do not develop disease even after immunization with the PLP 139-151 peptide, although inflammatory lesions can be seen in the brains of immunized mice. The appearance of lesions in the CNS of clinically normal mice following immunization with PLP 139-151 is consistent with data from mice protected with altered peptides that developed CNS lesions but were resistant to clinical disease

8 Feb 2002

10:20

112

AR

AR152-05.tex

AR152-05.SGM

LaTeX2e(2001/05/10)

P1: GSR

KUCHROO ET AL.

(74, 80); this argues for an active process by which this repertoire may lead to protection.

Annu. Rev. Immunol. 2002.20:101-123. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

The Activation of Naive Cross-Reactive Autoantigen Reactive T Cells The existence of two functionally different populations of PLP 139-151-reactive CD4+ T cells raises the question of how these cells could be differentially activated? This is important because if the two populations balance each others’ effects, then selective expansion of one or the other could influence disease susceptibility (Figure 1B). We have considered three possible explanations. The first is that because pathogenic and nonpathogenic TCRs use different contact residues, the two populations may react with PLP 139-151 with different avidities. Different levels of MHC and/or costimulation may then favor the development of different populations. The observation that anti-B7-1 antibody treatment shifts the balance of the pathogenic and nonpathogenic cells lends indirect support to this hypothesis. The second possibility is that the different populations are differentially sensitive to bystander activation, and the third possibility is that different populations are activated by different cross-reactive ligands. We believe that autoantigen reactive cells are always cross-reactive because of the overwhelming evidence that all TCRs can respond to a number of different peptides (81–83). Early work on antigen-specific T cell responses clearly demonstrated the potential for different ligands to activate the same TCR (84), but only more recently has it been appreciated that this is a common and important property of all T cells. The range of different peptide ligands that can activate a given TCR can be ordered in hierarchy based on their ability to stimulate (79). Some ligands may act as antagonists of T cell activation, and some may lead to anergy rather than activation (85). It is also clear from published data that TCRs can be activated by different peptides sharing little or no homology in their primary sequence (86). Estimates of how many peptides an individual TCR can interact with vary, but some are very large (104–107) (87). However, these estimates must be considered in terms of the likelihood that any particular TCR will encounter more than one peptide antigen for which it is specific. While this probability is not high for foreign peptides that are encountered at random, it may be much higher for those T cells for which cross-reactive ligands can be derived from self-peptides. It is not known with what frequency T cells in the periphery that have the potential to recognize self-antigen can become pathogenic effectors, but because such T cells can be identified in the repertoires of both humans and rodents, it is reasonable to assume they are present in significant numbers. There are two major non-exclusive hypotheses to explain how naive autoreactive T cells are activated. The first, bystander activation, postulates that autoreactive cells are activated in pro-inflammatory environments that occur in an organ when it is infected, for example with a virus. Under these circumstances, proinflammatory cytokines lead to increases in antigen presentation through the

8 Feb 2002

10:21

AR

AR152-05.tex

AR152-05.SGM

LaTeX2e(2001/05/10)

P1: GSR

Annu. Rev. Immunol. 2002.20:101-123. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

T CELL RESPONSE IN EAE

113

upregulation of MHC and costimulatory molecules that presumably reduce activation thresholds. Experimentally, bystander activation has been shown to promote the breakdown of self-tolerance in a TCR transgenic model of diabetes (88). In these experiments, infection with coxsackie B virus precipitates the onset of diabetes in mice transgenic for a diabetogenic TCR, but infection with a control virus, lymphocytic choriomeningitis virus, does not. Tests of cross-reactive activation were negative, leading the authors to conclude that activation of autopathogenic cells was the result of the pro-inflammatory microenvironment in the target organ. Another example of this mechanism is infection of the SJL mouse strain with Theiler’s murine encephalomyelitis virus (TMEV). TMEV induces a chronic CD4+ T cell–mediated demyelinating disease in the CNS. This chronic disease is due to the activation of PLP 139-151-reactive cells resulting from the de novo release of myelin antigens secondary to tissue destruction mediated by virusspecific T cells (89) and not due to cross-reactivation of PLP 139-151-reactive cells by TMEV. The second hypothesis is often termed molecular mimicry. This hypothesis proposes that activation of a T cell with a TCR that has the potential to interact with an autoantigen will change the cell from a naive nonpathogenic precursor into a pathogenic effector cell, which may be activated not only on exposure to the original pathogen, but also following encounter with autoantigen in the target organ. In support of this in both human and murine systems, it has been demonstrated that certain TCRs have the potential to respond to peptides derived from both autoantigens and pathogens (90–92). However, while many cross-reactive ligands have been identified, establishing clinical autoimmune disease by immunization with these ligands has in most cases proven to be difficult and often requires active infection and not simply immunization (93, 95). We have evaluated the role of cross-reactive ligands in the initiation of CNS autoimmunity in SJL mice using two approaches, by immunizing animals with analogs of PLP 139-151 identified by a database search of proteins expressed by foreign microorganisms (91), and by immunizing animals with APLs synthesized by mutating the primary TCR contact residue of PLP 139-151 (64, 74, 94). In both of these cases it is clear that immunization with a cross-reactive ligand expands a population of cells that reacts to the immunizing antigen and of which only a subset is cross-reactive with the autoantigen PLP 139-151. Immunizing SJL mice with peptides derived from murine hepatitis virus (MHV) and haemophilus influenzae type B (HAE) generated an immune response that cross-reacted with PLP 139-151. However, these mice did not develop EAE unless they were challenged with the autoantigen (91). Under these conditions we found that the threshold for disease induction had been reduced such that animals immunized with these peptides developed disease at a higher frequency than animals that had been immunized with non-cross-reactive control ligands. We postulated that lack of induction of disease by the HAE or MHV peptides may have been due to the lack of the inflammatory response normally induced by an infection (Figure 1B). To fulfill both these prerequisites of mimicry, together with nonspecific activation

8 Feb 2002

10:24

Annu. Rev. Immunol. 2002.20:101-123. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

114

AR

AR152-05.tex

AR152-05.SGM

LaTeX2e(2001/05/10)

P1: GSR

KUCHROO ET AL.

mediated by an infection, Miller and colleagues engineered the HAE mimicry peptide into a TMEV variant and showed that infection of SJL mice with the engineered virus induced rapid onset of disease without a need for any further immunization (95). Furthermore, infection with the engineered virus led to the induction of PLP 139-151-specific Th1 responses. These experiments are in contrast with the effects of immunization with an APL, Q144, which was synthesized by making a single amino acid substitution in the primary TCR contact of PLP 139-151. Pre-immunization with Q144 or co-immunization of Q144 with PLP 139-151 also generated an immune response that was cross-reactive with PLP 139-151. However, these cells were of a Th0/Th2 phenotype, and mice immunized with this ligand were resistant to the induction of EAE (64) (Figure 1B). Treatment of mice immunized with Q144 with anti-CTLA-4 antibody abrogated protection from disease induced with PLP 139-151 and decreased the frequency of cross-reactive IL-4 (and IL-2) secretors (95a). These data lead us to the conclusion that within the naive pool of PLP 139151-responsive T cells exist cells that can both promote and prevent EAE and that different cross-reactive ligands may selectively expand these different populations. Using APLs, our results show that we are not changing the function of the pathogenic repertoire, but expanding a distinct repertoire that is broadly cross-reactive and protective. Steinman and colleagues showed that some infectious organisms contain cross-reactive epitopes that, instead of enhancing autoimmunity by mimicry, may in fact act as APLs that induce protective cells and prevent development of autoimmunity (96). The cells that are expanded by the APLs or by infectious organisms have many properties similar to the CD4+CD25+ regulatory T cells that have recently been described and have been shown to maintain self-tolerance (97). Besides being generated in the thymus (98, 99), CD4+CD25+ cells may be induced in the periphery (100) by autoantigen and also by exposure to cross-reactive environmental antigens (discussed below). Both cytokine (IL-10 and TGF-β) and cell contact-dependent mechanisms of inhibition have been implicated, and these regulatory cells are functionally anergic and require activation through their TCR to become effectors. The hypothesis that exposure to environmental antigens may generate these regulatory T cells is consistent with the higher incidence of diabetes observed in NOD mice that have been reared in cleaner environments (101) and SJL mice that develop a higher frequency of PLP 139-151-reactive T cells in germ-free conditions (12).

The Activation and Regulation of Cross-Reactive Autoantigen Reactive Memory T Cells T cells bearing identical TCRs can have their function dramatically altered by the environment in which they are activated. This concept has been described theoretically as tuning (101a). The clearest example of this is the differentiation of naive transgenic T cells in vitro by different mixtures of cytokines. However, we believe

4 Feb 2002

11:21

AR

AR152-05.tex

AR152-05.SGM

LaTeX2e(2001/05/10)

P1: GSR

Annu. Rev. Immunol. 2002.20:101-123. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

T CELL RESPONSE IN EAE

115

that, especially for autoreactive cells, different effects on activation thresholds may also be important. Naive and memory cells with identical TCRs have different activation requirements. TCR transgenic memory cells, generated by activation in vitro and then transferred into unimmunized hosts, are less costimulation-dependent and have lower activation thresholds, as judged by the concentration of antigen required for their restimulation (102). For any T cell that has the potential to be activated by an autoantigen, this raises the possibility that memory cells originally activated by a pathogen-derived peptide might subsequently be restimulated by an autoantigen that was normally ignored by the naive cells. To investigate this, it is necessary to study TCRs that respond to both nonself- and self-peptides. In SJL mice we have analyzed T cell clones and transgenic T cells that are activated by the autoantigen PLP 139-151 and by nonself-ligands that are analogs of this peptide (103). Similarly, MBP Ac1-9 and its analogs have been studied by other groups in B10.PL (H-2u) mice (32). In this system, immunization with MBP Ac1-9 induces cells that are hyperresponsive to Ac1-9(4Tyr), but immunization with Ac1-9(4Tyr) induces a population of T cells that respond to Ac1-9(4Tyr) with the same dose response as the response of cells to Ac1-9. A detailed study concluded that Ac1-9 cells that are hyperresponsive to Ac1-9(4Tyr) are deleted when mice are immunized with this ligand (104); thus they do not survive to be detected in proliferation assays. In the PLP system, we have studied T cells from the 1B6 TCR transgenic mouse that respond to a number of ligands with a defined hierarchy. PLP 139-151, the autoantigen, is a weak ligand; Q144 is the cognate ligand; and a second peptide analog called L144 is a hyperstimulating ligand or superagonist (79). We find that activation of the naive transgenic T cells by equivalent concentrations of L144 and Q144 has very different effects on the subsequent phenotype of the lymphocytes that expand. Activation with the cognate ligand Q144 induces cells that retain the phenotype and dose response of the original T cell clone. On the other hand, activation with the superagonist L144 induces cells that continue to respond to L144, with half-maximal proliferation at higher antigen concentrations, but these cells are unable to respond to the autoantigen PLP 139-151 or to activation with antiCD3 antibody (M. Munder, L. B. Nicholson, and V. K. Kuchroo, submitted). These cells are now selectively tolerant to activation with the autoantigen while retaining their responses to L144. In this case the cells do not undergo activation-induced cell death, but biochemically they appear to be anergic. Selectively tolerant 1B6 T cells require higher concentrations of L144 for optimal activation than do the naive cells in vitro, but they have become refractory to activation by autoantigen. Knowing that T cells are generally cross-reactive, such a response in vivo would decrease degeneracy and improve the specificity and fidelity of the memory immune response, by preserving the useful nonself response while eliminating the potentially injurious responses to self. It may be that apoptosis in the MBP system and selective PLP tolerance in the system lie along a spectrum of responses that help preserve self-tolerance without sacrificing a broad peripheral immune repertoire necessary for effective responses to pathogens.

8 Feb 2002

10:24

116

AR

AR152-05.tex

AR152-05.SGM

LaTeX2e(2001/05/10)

P1: GSR

KUCHROO ET AL.

Annu. Rev. Immunol. 2002.20:101-123. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

ACTIVATION THRESHOLDS, HIERARCHIES, AND INDUCTION OF PATHOGENIC, PROTECTIVE, AND REGULATORY MEMORY The picture of the T cell that begins to emerge from these studies is of a cell that uses a rather degenerate recognition apparatus. However, regulation of activation thresholds in both the thymus and the periphery allows them to achieve functional specificity in most immune responses. In the thymus, autoreactive T cells survive negative selection below a certain threshold. For example, in I-Au mice, cells that can be activated by MBP Ac1-11 peptide escape deletion and become the dominant MBP-reactive population in the periphery because the weak binding of Ac1-11 and I-Au leads to low-avidity interactions with TCRs. In the periphery, activation with stronger antigens can blunt responses to weaker ligands, leading to an increase in specificity at the expense of sensitivity. Both in vivo and in vitro studies of responses to single foreign peptides are therefore confounded by the ability of T cells to adapt their activation thresholds to the immunizing antigen. What may be happening to the responsiveness to self-antigen may be masked by this property. We and others have approached this by identifying TCRs with specific autoreactivities and then studying their response to self-antigen following activation with foreign antigen. This remains a limited approach because it does not deal with the ability of whole populations of antigen-reactive T cells, responding to a single peptide, to encompass a broad range of different responses to a cross-reactive autoantigen. However, analyses like this, at the single TCR level, allow us to discern the functional response of a single T cell to multiple ligands (Figure 2). These studies show that even a single TCR can have a broad range of functional responses when activated with different ligands. On the basis of these studies we postulate that although it is common for one or more of the antigens recognized in the hierarchy of a TCR to be an autoantigen, changes in activation threshold do not usually favor the development of autoreactivity. Because of thymic selection, on average TCRs have a lower avidity for self-antigens than for foreign antigens. During activation of a T lymphocyte by an infectious foreign antigen, protective memory is usually generated, and these cells are kept in the peripheral pool where they will mediate protection against future infections by the same or similar infectious agents. When the nonself-antigen is much higher in the hierarchy than self-antigen, activation can result in selection and/or tuning of the activation threshold, lowering the responsiveness of the resultant memory cells to the self-antigen. Such memory cells will show selective nonresponsiveness against the self-antigen, yet retain responses to the nonselfantigen; this we call protective/regulatory memory against self. This may be one of the mechanisms by which CD4+CD25+ regulatory memory T cells against self-antigens are generated in the peripheral T cell repertoire; self-reactive T cells that are activated by the superagonist nonself-antigen become anergic to activation with self- or with anti-CD3 antibody. If the nonself-antigen is lower in hierarchy than the self, the memory cells generated will be more sensitive to activation with

8 Feb 2002

10:26

AR

AR152-05.tex

AR152-05.SGM

LaTeX2e(2001/05/10)

P1: GSR

Annu. Rev. Immunol. 2002.20:101-123. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

T CELL RESPONSE IN EAE

117

the self-antigen. The resulting memory cells will show increased responsiveness and a lower threshold for activation by the self-antigen and will have the potential to induce autoimmunity. This we designate as pathogenic memory for self (Figure 2). Knowing that TCRs are generally cross-reactive, it is obvious that the effects of thymic selection, in establishing autoreactivities lower in the hierarchy than responses to foreign antigens, are critical to self-nonself discrimination. The mechanism of induction of selective tolerance would explain why expansion by nonself-antigens does not usually result in the induction of an autoimmune disease. Furthermore, this also suggests a mechanism by which cross-reactive memory T cells might mediate responses to nonself yet maintain tolerance to self and prevent development of an autoimmune disease. Understanding how these pathogenic and nonpathogenic self-reactive cells are regulated in health and activated in disease remains an important challenge for fuller understanding of the pathogenesis and development of nontoxic antigenbased therapies for organ-specific autoimmune disease. ACKNOWLEDGMENTS The work from our laboratory discussed in this paper was supported by grants from the National Institutes of Health and the National Multiple Sclerosis Society. We are grateful to all our colleagues for their contributions. We would also like to thank Sara Abromson-Leeman of Harvard Medical School for reading the manuscript and Stephen Miller of Northwestern University Medical School for sharing data before publication. Visit the Annual Reviews home page at www.annualreviews.org

LITERATURE CITED 1. Whitehouse DJ, Whitehouse MW, Pearson CM. 1969. Passive transfer of adjuvant-induced arthritis and allergic encephalomyelitis in rats using thoracic duct lymphocytes. Nature 224:1322 2. Alvord EC, Kies MW, Suckling AJ. 1984. Experimental Allergic Encephalomyelitis: A Useful Model for Multiple Sclerosis. New York: Liss 3. Zamvil SS, Steinman L. 1990. The T lymphocyte in experimental allergic encephalomyelitis. Annu. Rev. Immunol. 8: 579–621 4. Whitham RH, Bourdette DN, Hashim GA, Herndon RM, Ilg RC, Vandenbark

AA, Offner H. 1991. Lymphocytes from SJL/J mice immunized with spinal cord respond selectively to a peptide of proteolipid protein and transfer relapsing demyelinating experimental autoimmune encephalomyelitis. J. Immunol. 146:101– 7 5. Kuchroo VK, Sobel RA, Laning JC, Martin CA, Greenfield E, Dorf ME, Lees MB. 1992. Experimental allergic encephalomyelitis mediated by cloned T cells specific for a synthetic peptide of myelin proteolipid protein. Fine specificity and T cell receptor Vβ usage. J. Immunol. 148:3776–82

4 Feb 2002

11:21

Annu. Rev. Immunol. 2002.20:101-123. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

118

AR

AR152-05.tex

AR152-05.SGM

LaTeX2e(2001/05/10)

P1: GSR

KUCHROO ET AL.

6. Baron JL, Madri JA, Ruddle NH, Hashim G, Janeway CA. 1993. Surface expression of α4 integrin by CD4 T cells is required for their entry into brain parenchyma. J. Exp. Med. 177:57–68 7. Chang TT, Jabs C, Sobel RA, Kuchroo VK, Sharpe AH. 1999. Studies in B7deficient mice reveal a critical role for B7 costimulation in both induction and effector phases of experimental autoimmune encephalomyelitis. J. Exp. Med. 190:733– 40 8. Williams KC, Ulvestad E, Hickey WF. 1994. Immunology of multiple sclerosis. Neuroscience 2:229–45 9. Sobel RA, Kuchroo VK. 1992. The immunopathology of acute allergic encephalomyelitis induced with myelin proteolipid protein. J. Immunol. 149:1444–51 10. Zhang J, Markovic S, Lacet B, Raus J, Weiner HL, Hafler DA. 1994. Increased frequency of interleukin 2-responsive cells specific for myelin basic protein in peripheral blood and cerebrospinal fluid of patients with multiple sclerosis. J. Exp. Med. 179:973–84 11. Markovic-Plese S, Fukaura H, Zhang J, Al-Sabbagh A, Southwood S, Sette A, Kuchroo VK, Hafler DA. 1995. T cell recognition of immunodominant and cryptic proteolipid protein epitopes in humans. J. Immunol. 155:982–92 12. Anderson AC, Nicholson LB, Legge KL, Turchin V, Zaghouani H, Kuchroo VK. 2000. High frequency of autoreactive myelin proteolipid protein-specific T cells in the periphery of naive mice: mechanisms of selection of the self-reactive repertoire. J. Exp. Med. 191:761–70 13. Waldner H-P, Whitters MJ, Sobel RA, Collins M, Kuchroo VK. 2000. Fulminant spontaneous autoimmunity of the central nervous system in myelin proteolipid protein specific T cell receptor transgenic mice. Proc. Natl. Acad. Sci. USA 97:3412–17 14. Nossal GJV. 1994. Negative selection of lymphocytes. Cell 76:229–39

15. von Boehmer H. 1994. Positive selection of lymphocytes. Cell 76:219–28 16. Williams KC, Hickey WF. 1995. Traffic of hematogenous cells through the central nervous system. Curr. Top. Microb. Immunol. 202:221–45 17. Weller RO, Engelhardt B, Phillips MJ. 1996. Lymphocyte targeting of the central nervous system: a review of afferent and efferent CNS-immune pathways. Brain Pathol. 6:275–88 18. Pribyl TM, Campagnoni CW, Kampf K, Kashima T, Handley VW, McMahon J, Campagnoni AT. 1993. The human myelin basic protein gene is included within a 179-kilobase transcription unit: expression in the immune and central nervous systems. Proc. Natl. Acad. Sci. USA 90:10695–99 19. Pribyl TM, Campagnoni CW, Kampf K, Kashima T, Handley VW, McMahon J, Campagnoni AT. 1996. Expression of the myelin proteolipid protein gene in the human fetal thymus. J. Neuroimmunol. 67:125–30 20. Mathisen PM, Pease S, Garvey J, Hood L, Readhead C. 1993. Identification of an embryonic isoform of myelin basic protein that is expressed widely in the mouse embryo. Proc. Natl. Acad. Sci. USA 90:10125–29 21. Fritz RB, Zhao M-L. 1996. Thymic expression of myelin basic protein (MBP). J. Immunol. 157:5249–53 22. Voskuhl RR. 1998. Myelin protein expression in lymphoid tissues: implications for peripheral tolerance. Immunol. Rev. 164:81–92 23. Spitzweg C, Joba W, Heufelder AE. 1999. Expression of thyroid-related genes in human thymus. Thyroid 9:133–41 24. Sospedra M, Ferrer-Francesch X, Domiguez O, Juan M, Foz-Sala M, PujolBorrell R. 1998. Transcription of a broad range of self-antigens in human thymus suggests a role for central mechanisms in tolerance toward peripheral antigens. J. Immunol. 161:5918–29

4 Feb 2002

11:21

AR

AR152-05.tex

AR152-05.SGM

LaTeX2e(2001/05/10)

P1: GSR

Annu. Rev. Immunol. 2002.20:101-123. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

T CELL RESPONSE IN EAE 25. Charukamnoetkanok P, Fukushima A, Whitcup SM, Gery I, Egwuagu CE. 1998. Expression of ocular autoantigens in the mouse thymus. Curr. Eye Res. 17:788– 92 26. Heath VL, Moore NC, Parnell SM, Mason DW. 1998. Intrathymic expression of genes involved in organ specific autoimmune disease. J. Autoimmunol. 11:309– 18 27. Egwuagu CE, Charukamnoetkanok P, Gery I. 1997. Thymic expression of autoantigens correlates with resistance to autoimmune disease. J. Immunol. 159:3109– 12 28. Targoni OS, Lehmann PV. 1998. Endogenous myelin basic protein inactivates the high avidity T cell repertoire. J. Exp. Med. 187:2055–63 29. Yoshizawa I, Bronson R, Ben-Nun A, Richert JR, Dorf ME, Abromson-Leeman S. 1998. Differential recognition of MBP epitopes in BALB/c mice determines the site of inflammatory disease induction. J. Neuroimmunol. 89:73–82 30. Klein L, Klugmann M, Nave KA, Tuohy VK, Kyewski B. 2000. Shaping of the autoreactive T-cell repertoire by a splice variant of self protein expressed in thymic epithelial cells. Nat. Med. 6:56–61 31. Correia-Neves M, Waltzinger C, Mathis D, Benoist C. 2001. The shaping of the T cell repertoire. Immunity 14:21–32 32. Kumar V, Bhardwaj V, Soares L, Alexander J, Sette A, Sercarz E. 1995. Major histocompatibility complex binding affinity of an antigenic determinant is crucial for the differential secretion of interleukin 4/5 or interferon gamma by T cells. Proc. Natl. Acad. Sci. USA 92:9510–14 33. Harrington CJ, Paez A, Hunkapiller T, Mannikko V, Brabb T, Ahearn M, Beeson C, Goverman J. 1998. Differential tolerance is induced in T cells recognizing distinct epitopes of myelin basic protein. Immunity 8:571–80 34. Liu GY, Fairchild PJ, Smith RM, Prowie JR, Kioussis D, Wraith DC. 1995. Low

34a.

35.

36.

37.

38.

39.

40.

41.

42.

119

avidity recognition of self-antigen by T cells permits escape from central tolerance. Immunity 3:407–15 Pribyl TM, Campagnoni CW, Kampf K, Kashima T, Handley VW, McMahon J, Campagnoni AT. 1993. The human myelin basic protein gene is included within a 179-kilobase transcription unit: expression in the immune and central nervous systems. Proc. Natl. Acad. Sci. USA 90:10695–99 Huseby E, Sather B, Huseby PB, Goverman J. 2001. Age-dependent T cell tolerance and autoimmunity to myelin basic protein. Immunity 14:471–81 Wicker LS, Todd JA, Peterson LB. 1995. Genetic control of autoimmune diabetes in the NOD mouse. Annu. Rev. Immunol. 13:179–200 Ridgway WM, Fasso M, Lanctot A, Garvey C, Fathman CG. 1996. Breaking selftolerance in nonobese diabetic mice. J. Exp. Med. 183:1657–62 Kanagawa O, Martin SM, Vaupel BA, Carrasco-Marin E, Unanue ER. 1998. Autoreactivity of T cells from nonobese diabetic mice: an I-Ag7-dependent reaction. Proc. Natl. Acad. Sci. USA 95:1721–24 Ridgway WM, Ito H, Fasso M, Yu C, Fathman CG. 1998. Analysis of the role of variation of major histocompatibility complex class II expression on nonobese diabetic (NOD) peripheral T cell response. J. Exp. Med. 188:2267–75 Carrasco-Marin E, Shimizu J, Kanagawa O, Unanue ER. 1996. The class II MHC I-Ag7 molecules from non-obese diabetic mice are poor peptide binders. J. Immunol. 156:450–58 Reizis B, Eisenstein M, Bockova J, Konen-Waisman S, Mor F, Elias D, Cohen IR. 1997. Molecular characterization of the diabetes-associated mouse MHC class II protein, I-Ag7. Int. Immunol. 9:43–51 Peterson M, Sant AJ. 1998. The inability of the nonobese diabetic class II molecule to form stable peptide complexes does not

4 Feb 2002

11:21

120

43.

Annu. Rev. Immunol. 2002.20:101-123. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

44.

45.

46.

47.

48.

49.

50. 51. 52.

53.

AR

AR152-05.tex

AR152-05.SGM

LaTeX2e(2001/05/10)

P1: GSR

KUCHROO ET AL. reflect a failure to interact productively with DM. J. Immunol. 161:2961–67 Hausmann DH, Yu B, Hausmann S, Wucherpfennig KW. 1999. pH-dependent peptide binding properties of the type I diabetes-associated I-Ag7 molecule: rapid release of CLIP at an endosomal pH. J. Exp. Med. 189:1723–34 Ridgway WM, Fasso M, Fathman CG. 1999. A new look at MHC and autoimmune disease. Science 284:749 Encinas JA, Lees MB, Sobel RA, Symonowicz C, Greer JA, Shovlin CL, Weiner HL, Seidman CE, Seidman JG, Kuchroo VK. 1996. Genetic analysis of susceptibility to experimental autoimmune encephalomyelitis in a cross between SJL/J and B10.S mice. J. Immunol. 157:2186–92 Powrie F, Coffman RL. 1993. Cytokine regulation of T-cell function: potential for therapeutic intervention. Immunol. Today 14:270–74 Mosmann TR, Cherwinski H, Bond MW, Giedlin MA, Coffman RL. 1986. Two types of murine helper T cell clones. J. Immunol. 136:2348–57 Abbas AK, Murphy KM, Sher A. 1996. Functional diversity of helper T lymphocytes. Nature 383:787–93 Mosmann TR, Coffman RL. 1989. Th1 and Th2 cells: different patterns of lymphokine secretion lead to different functional properties. Annu. Rev. Immunol. 7:145–73 Paul WE, Seder RA. 1994. Lymphocyte responses and cytokines. Cell 76:241–51 Swain SL. 1993. Polarized patterns of cytokine secretion. Curr. Biol. 3:115–17 Hsieh C, Macatonia SE, Tripp CS, Wolf SF, Ogarra A, Murphy KM. 1993. Development of Th1 CD4+ T cells through IL12 produced by listeria-induced macrophages. Science 260:547–49 Miller SD, Karpus WJ. 1994. The immunopathogenesis and regulation of Tcell-mediated demyelinating diseases. Immunol. Today 15:356–61

54. Ben-Nun A, Cohen I. 1982. Experimental allergic encephalomyelitis mediated by cloned T cells specific for a synthetic peptide of myelin proteolipid protein. Fine specificity and T cell receptor Vβ usage. J. Immunol. 129:303–8 55. Zamvil S, Nelson P, Trotter J, Mitchell D, Knobler R, Fritz R, Steinman L. 1985. T-cell clones specific for myelin basic protein induce chronic relapsing paralysis and demyelination. Nature 317:355– 58 56. Khoury SJ, Hancock WW, Weiner HL. 1992. Oral tolerance to myelin basic protein and natural recovery from experimental autoimmune encephalomyelitis are associated with downregulation of inflammatory cytokines and differential upregulation of transforming growth factor beta, interleukin 4, and prostaglandin E expression in the brain. J. Exp. Med. 176:1355– 64 57. Merrill JE, Kono DH, Clayton J, Ando DG, Hinton DR, Hofman FM. 1992. Inflammatory leukocytes and cytokines in the peptide-induced disease of experimental allergic encephalomyelitis in SJL and B10.PL mice. Proc. Natl. Acad. Sci. USA 89:574–78 58. Leonard JP, Waldburger KE, Goldman SJ. 1995. Prevention of experimental autoimmune encephalomyelitis by antibodies against interleukin 12. J. Exp. Med. 181:381–86 59. Karpus WJ, Gould KE, Swanborg RH. 1992. CD4+ suppressor cells of autoimmune encephalomyelitis respond to T cell receptor-associated determinants on effector cells by interleukin-4 secretion. Eur. J. Immunol. 22:1757–63 60. Chen Y, Kuchroo VK, Inobe J, Hafler DA, Weiner HL. 1994. Regulatory T cell clones induced by oral tolerance: suppression of autoimmune encephalomyelitis. Science 265:1237–40 61. Kennedy MK, Torrance DS, Picha KS, Mohler KM. 1992. Analysis of cytokine mRNA expression in the central nervous

4 Feb 2002

11:21

AR

AR152-05.tex

AR152-05.SGM

LaTeX2e(2001/05/10)

P1: GSR

T CELL RESPONSE IN EAE

Annu. Rev. Immunol. 2002.20:101-123. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

62.

63.

64.

65.

66.

67.

68.

69.

70.

system of mice with experimental autoimmune encephalomyelitis reveals that IL10 mRNA expression correlates with recovery. J. Immunol. 149:2496–505 Racke MK, Bonomo A, Scott DE, Cannella B, Levine A, Raine CS, Shevach EM, Rocken M. 1994. Cytokine-induced immune deviation as a therapy for inflammatory autoimmune disease. J. Exp. Med. 180:1961–66 Kuchroo VK, Prabhu Das M, Brown JA, Ranger AM, Zamvil SS, Sobel RA, Weiner HL, Nabavi N, Glimcher LH. 1995. B7-1 and B7-2 costimulatory molecules activate differentially the Th1/Th2 developmental pathways: application to autoimmune disease therapy. Cell 80:707–8 Nicholson LB, Greer JM, Sobel RA, Lees MA, Kuchroo VK. 1995. An altered peptide ligand mediates immune deviation and prevents experimental autoimmune encephalomyelitis. Immunity 3:397–405 Bettelli E, Das MP, Howard ED, Weiner HL, Sobel RA, Kuchroo VK. 1998. IL-10 is critical in the regulation of autoimmune encephalomyelitis as demonstrated by studies of IL-10- and IL-4deficient and transgenic mice. J. Immunol. 161:3299–306 Segal BM, Dwyer BK, Shevach EM. 1998. An interleukin (IL)-10/IL-12 immunoregulatory circuit controls susceptibility to autoimmune disease. J. Exp. Med. 187:537–46 Cua DJ, Groux H, Hinton DR, Stohlman SA, Coffman RL. 1999. Transgenic interleukin 10 prevents induction of experimental autoimmune encephalomyelitis. J. Exp. Med. 189:1005–10 Segal BM, Shevach EM. 1996. IL-12 unmasks latent autoimmune disease in resistant mice. J. Exp. Med. 184:771–75 Segal BM, Klinman DM, Shevach EM. 1997. Microbial products induce autoimmune disease by an IL-12-dependent pathway. J. Immunol. 158:5087–90 Encinas JA, Wicker LS, Peterson LB,

71.

72.

73.

74.

75.

76.

77.

121

Mukasa A, Teuscher C, Sobel R, Weiner HL, Seidman CE, Seidman JG, Kuchroo VK. 1999. QTL influencing autoimmune diabetes and encephalomyelitis map to a 0.15-cM region containing Il2. Nat. Genet. 21:158–60 Maron R, Hancock WW, Slavin A, Hattori M, Kuchroo V, Weiner HL. 1999. Genetic susceptibility or resistance to autoimmune encephalomyelitis in MHC congenic mice is associated with differential production of pro- and antiinflammatory cytokines. Int. Immunol. 11:1573–80 Khoruts A, Miller SD, Jenkins MK. 1995. Neuroantigen-specific Th2 cells are inefficient suppressors of experimental autoimmune encephalomyelitis induced by effector Th1 cells. J. Immunol. 155:5011– 17 Lafaille JJ, Van de Keere F, Hsu AL, Baron JL, Haas W, Raine CS, Tonegawa S. 1997. Myelin basic protein-specific T helper 2 (Th2) cells cause experimental autoimmune encephalomyelitis in immunodeficient hosts rather than protect them from the disease. J. Exp. Med. 186: 307–12 Kuchroo VK, Greer JM, Kaul D, Ishioka G, Franco A, Sette A, Sobel RA, Lees MB. 1994. A single TCR antagonist peptide inhibits experimental allergic encephalomyelitis mediated by a diverse T cell repertoire. J. Immunol. 153:3326–36 Franco A, Southwood S, Arrhenius T, Kuchroo VK, Grey HM, Sette A, Ishioka GY. 1994. T cell receptor antagonist peptides are highly effective inhibitors of experimental allergic encephalomyelitis. Eur. J. Immunol. 24:940–46 McRae BL, Miller SD. 1994. Fine specificity of CD4+ T cell responses to the dominant encephalitogenic PLP 139-151 peptide in SJL/J mice. Neurochem. Res. 19:997–1004 van der Veen RC, Chen P, McMillan M. 1995. Myelin proteolipid protein-induced Th1 and Th2 clones express TCR with

4 Feb 2002

11:21

122

78.

Annu. Rev. Immunol. 2002.20:101-123. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

79.

80.

81.

82.

83.

84.

85.

86.

87.

AR

AR152-05.tex

AR152-05.SGM

LaTeX2e(2001/05/10)

P1: GSR

KUCHROO ET AL. similar fine specificity for peptide and CDR3 homology despite diverse Vβ usage. Cell. Immunol. 166:291–95 Prabhu Das M, Nicholson LB, Greer JM, Kuchroo VK. 1997. Auto-pathogenic Th1 and protective Th2 clones differ in their recognition of the autoantigenic peptide of myelin proteolipid protein. J. Exp. Med. 186:867–76 Nicholson LB, Waldner H-P, Carrizosa A, Sette A, Collins M, Kuchroo VK. 1998. Heteroclitic proliferative responses and changes in cytokine profile induced by altered peptides: implications for autoimmunity. Proc. Natl. Acad. Sci. USA 95:264–69 Santambrogio L, Lees MB, Sobel RA. 1998. Altered peptide ligand modulation of experimental allergic encephalomyelitis: immune responses within the CNS. J. Neuroimmunol. 81:1–13 Germain RN, Levine EL, Madrenas J. 1995. The T-cell receptor as a diverse signal transduction machine. Immunologist 3:113–21 Kersh GJ, Allen PM. 1996. Essential flexibility in the T-cell antigen recognition of antigen. Nature 380:495–98 Goldrath AW, Bevan MJ. 1999. Selecting and maintaining a diverse T-cell repertoire. Nature 402:255–62 Hedrick SM, Matis LA, Hecht TT, Samelson LE, Longo DL, Heber-Katz E, Schwartz RH. 1982. The fine specificity of antigen and Ia determinant recognition by T cell hybridoma clones specific for pigeon cytochrome c. Cell 30:141–52 Nicholson LB, Kuchroo VK. 1997. T cell recognition of self and altered self antigens. Crit. Rev. Immunol. 17:449–62 Evavold BD, Sloan-Lancaster J, Wilson KJ, Rothbard JB, Allen PM. 1995. Specific T cell recognition of minimally homologous peptides: evidence for multiple endogenous ligands. Immunity 2:655–63 Mason D. 1998. A very high level of crossreactivity is an essential feature of the Tcell receptor. Immunol. Today 19:395–404

88. Horwitz MS, Bradley LM, Harbertson J, Krahl T, Lee J, Sarvetnick N. 1998. Diabetes induced by Coxsackie virus: initiation by bystander damage and not molecular mimicry. Nat. Med. 4:781–85 89. Miller SD, Vanderlugt CL, Begolka WS, Pao W, Yauch RL, Neville KL, KatzLevy Y, Carrizosa A, Kim BS. 1997. Persistent infection with Theiler’s virus leads to CNS autoimmunity via epitope spreading. Nat. Med. 3:1133–36 90. Fujinami RS, Oldstone MB. 1985. Amino acid homology between the encephalitogenic site of myelin basic protein and virus: mechanism for autoimmunity. Science 230:1043–45 91. Carrizosa AM, Nicholson LB, Farzan M, Southwood S, Sette A, Sobel RA, Kuchroo VK. 1998. Expansion by self antigen is necessary for the induction of experimental autoimmune encephalomyelitis by T cells primed with a crossreactive environmental antigen. J. Immunol. 161:3307–14 92. Wucherpfennig KW, Strominger JL. 1995. Molecular mimicry in T cellmediated autoimmunity: viral peptides activate human T cell clones specific for myelin basic protein. Cell 80:695–705 93. Zhao ZS, Granucci F, Yeh L, Schaffer PA, Cantor H. 1998. Molecular mimicry by herpes simplex virus-type 1: autoimmune disease after viral infection. Science 279:1344–47 94. Nicholson LB, Murtaza A, Hafler BP, Sette A, Kuchroo VK. 1997. A T cell receptor antagonist peptide prevents autoimmune encephalomyelitis induced with multiple myelin antigens: evidence for bystander suppression. Proc. Natl. Acad. Sci. USA 94:9279–84 95. Olson JK, Croxford JL, Calenoff MA, Dal Canto MC, Miller SD. 2001. A virus-induced molecular mimicry model of multiple sclerosis. J. Clin. Invest. 108:311–18 95a. Kuhns MS, Epshteyn V, Sobel RA, Allison JP. 2000. Cytotoxic T lymphocyte

4 Feb 2002

11:21

AR

AR152-05.tex

AR152-05.SGM

LaTeX2e(2001/05/10)

P1: GSR

Annu. Rev. Immunol. 2002.20:101-123. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

T CELL RESPONSE IN EAE antigen-4 (CTLA-4) regulates the size, reactivity, and function of a primed pool of CD4+ cells. Proc. Natl. Acad. Sci. USA 97:12711–16 96. Ruiz PJ, Garren H, Hirschberg DL, Langer-Gould AM, Levite M, Karpuj MV, Southwood S, Sette A, Conlon P, Steinman L. 1999. Microbial epitopes act as altered peptide ligands to prevent experimental autoimmune encephalomyelitis. J. Exp. Med. 189:1275–84 97. Sakaguchi S. 2001. Policing the regulators. Nat. Immunol. 2:283–84 98. Saoudi A, Seddon B, Fowell D, Mason D. 1996. The thymus contains a high frequency of cells that prevent autoimmune diabetes on transfer into prediabetic recipients. J. Exp. Med. 184:2393–98 99. Itoh M, Takahashi T, Sakaguchi N, Kuniyasu Y, Shimizu J, Otsuka F, Sakaguchi S. 1999. Thymus and autoimmunity: production of CD25+ CD4+ naturally anergic and suppressive T cells as a key function of the thymus in maintaining immunologic self-tolerance. J. Immunol. 162:5317–26 100. Seddon B, Mason D. 1999. Peripheral

101.

101a.

102.

103.

104.

123

autoantigen induces regulatory T cells that prevent autoimmunity. J. Exp. Med. 189:877–82 Leiter EH, Serreze DV, Prochazka M. 1990. The genetics and epidemiology of diabetes in NOD mice. Immunol. Today 11:147–49 Grossman Z, Paul WE. 1992. Adaptive cellular interactions in the immune system: the tunable activation threshold and the significance of subthreshold responses. Proc. Natl. Acad. Sci. USA 89:10365–69 London CA, Lodge MP, Abbas AK. 2000. Functional responses and costimulator dependence of memory CD4+ T cells. J. Immunol. 164:265–72 Nicholson LB, Anderson AC, Kuchroo VK. 2000. Tuning T cell activation threshold and effector function with cross-reactive peptide ligands. Int. Immunol. 12:205–13 Anderton SM, Radu CG, Lowrey PA, Ward ES, Wraith DC. 2001. Negative selection during the peripheral immune response to antigen. J. Exp. Med. 193:1– 11

8:17 AR AR152-05-COLOR.tex AR152-05-COLOR.SGM

Figure 1 Two PLP 139-151 specific repertoires. (A) Two repertoires, which differ in the fine specificity of their recognition of PLP 139-151, are used by pathogenic and non-pathogenic T cells. (B) The two repertoires balance each other in healthy individuals, but expansion of one or the other can result in disease or protection.

Annu. Rev. Immunol. 2002.20:101-123. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

6 Feb 2002 LaTeX2e(2002/01/18) P1: GDL

Annu. Rev. Immunol. 2002.20:101-123. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

6 Feb 2002

8:17

AR

AR152-05-COLOR.tex

AR152-05-COLOR.SGM

LaTeX2e(2002/01/18)

P1: GDL

Figure 2 Exposure to cross-reactive non-self antigen can induce protective, regulatory or pathogenic memory cells. Activation with a strong (superagonist) non-selfligand lowers the responses of the memory population to self and leads to the development of protective memory. Activation with a weak cross-reactive non-self-antigen will increase responsiveness and leads to pathogenic memory which has the potential to induce autoimmunity.

P1: FRK

February 1, 2002

12:47

Annual Reviews

AR152-FM

Annual Review of Immunology Volume 20, 2002

Annu. Rev. Immunol. 2002.20:101-123. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

CONTENTS FRONTISPIECE, Charles A. Janeway, Jr. A TRIP THROUGH MY LIFE WITH AN IMMUNOLOGICAL THEME, Charles A. Janeway, Jr.

xiv 1

THE B7 FAMILY OF LIGANDS AND ITS RECEPTORS: NEW PATHWAYS FOR COSTIMULATION AND INHIBITION OF IMMUNE RESPONSES, Beatriz M. Carreno and Mary Collins

29

MAP KINASES IN THE IMMUNE RESPONSE, Chen Dong, Roger J. Davis, and Richard A. Flavell

PROSPECTS FOR VACCINE PROTECTION AGAINST HIV-1 INFECTION AND AIDS, Norman L. Letvin, Dan H. Barouch, and David C. Montefiori T CELL RESPONSE IN EXPERIMENTAL AUTOIMMUNE ENCEPHALOMYELITIS (EAE): ROLE OF SELF AND CROSS-REACTIVE ANTIGENS IN SHAPING, TUNING, AND REGULATING THE AUTOPATHOGENIC T CELL REPERTOIRE, Vijay K. Kuchroo, Ana C. Anderson, Hanspeter Waldner, Markus Munder, Estelle Bettelli, and Lindsay B. Nicholson

55 73

101

NEUROENDOCRINE REGULATION OF IMMUNITY, Jeanette I. Webster, Leonardo Tonelli, and Esther M. Sternberg

125

MOLECULAR MECHANISM OF CLASS SWITCH RECOMBINATION: LINKAGE WITH SOMATIC HYPERMUTATION, Tasuku Honjo, Kazuo Kinoshita, and Masamichi Muramatsu

165

INNATE IMMUNE RECOGNITION, Charles A. Janeway, Jr. and Ruslan Medzhitov

KIR: DIVERSE, RAPIDLY EVOLVING RECEPTORS OF INNATE AND ADAPTIVE IMMUNITY, Carlos Vilches and Peter Parham ORIGINS AND FUNCTIONS OF B-1 CELLS WITH NOTES ON THE ROLE OF CD5, Robert Berland and Henry H. Wortis E PROTEIN FUNCTION IN LYMPHOCYTE DEVELOPMENT, Melanie W. Quong, William J. Romanow, and Cornelis Murre

197 217 253 301

LYMPHOCYTE-MEDIATED CYTOTOXICITY, John H. Russell and Timothy J. Ley

SIGNAL TRANSDUCTION MEDIATED BY THE T CELL ANTIGEN x

323

P1: FRK

Annu. Rev. Immunol. 2002.20:101-123. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

February 1, 2002

12:47

Annual Reviews

AR152-FM

CONTENTS

xi

RECEPTOR: THE ROLE OF ADAPTER PROTEINS, Lawrence E. Samelson INTERACTION OF HEAT SHOCK PROTEINS WITH PEPTIDES AND ANTIGEN PRESENTING CELLS: CHAPERONING OF THE INNATE AND ADAPTIVE IMMUNE RESPONSES, Pramod Srivastava CHROMATIN STRUCTURE AND GENE REGULATION IN THE IMMUNE SYSTEM, Stephen T. Smale and Amanda G. Fisher PRODUCING NATURE’S GENE-CHIPS: THE GENERATION OF PEPTIDES FOR DISPLAY BY MHC CLASS I MOLECULES, Nilabh Shastri, Susan

371

Schwab, and Thomas Serwold

395 427

463

THE IMMUNOLOGY OF MUCOSAL MODELS OF INFLAMMATION, Warren Strober, Ivan J. Fuss, and Richard S. Blumberg T CELL MEMORY, Jonathan Sprent and Charles D. Surh

GENETIC DISSECTION OF IMMUNITY TO MYCOBACTERIA: THE HUMAN MODEL, Jean-Laurent Casanova and Laurent Abel ANTIGEN PRESENTATION AND T CELL STIMULATION BY DENDRITIC CELLS, Pierre Guermonprez, Jenny Valladeau, Laurence Zitvogel, Clotilde Th´ery, and Sebastian Amigorena

495 551 581

621

NEGATIVE REGULATION OF IMMUNORECEPTOR SIGNALING, Andr´e Veillette, Sylvain Latour, and Dominique Davidson

669

CPG MOTIFS IN BACTERIAL DNA AND THEIR IMMUNE EFFECTS, Arthur M. Krieg

PROTEIN KINASE Cθ

709 IN

T CELL ACTIVATION, Noah Isakov and Amnon

Altman

RANK-L AND RANK: T CELLS, BONE LOSS, AND MAMMALIAN EVOLUTION, Lars E. Theill, William J. Boyle, and Josef M. Penninger PHAGOCYTOSIS OF MICROBES: COMPLEXITY IN ACTION, David M. Underhill and Adrian Ozinsky

761 795 825

STRUCTURE AND FUNCTION OF NATURAL KILLER CELL RECEPTORS: MULTIPLE MOLECULAR SOLUTIONS TO SELF, NONSELF DISCRIMINATION, Kannan Natarajan, Nazzareno Dimasi, Jian Wang, Roy A. Mariuzza, and David H. Margulies

853

INDEXES Subject Index Cumulative Index of Contributing Authors, Volumes 1–20 Cumulative Index of Chapter Titles, Volumes 1–20

ERRATA An online log of corrections to Annual Review of Immunology chapters (if any, 1997 to the present) may be found at http://immunol.annualreviews.org/

887 915 925

8 Feb 2002

9:39

AR

AR152-06.tex

AR152-06.SGM

LaTeX2e(2001/05/10) P1: GJC 10.1146/annurev.immunol.20.082401.104914

Annu. Rev. Immunol. 2002. 20:125–63 DOI: 10.1146/annurev.immunol.20.082401.104914

NEUROENDOCRINE REGULATION OF IMMUNITY∗

Annu. Rev. Immunol. 2002.20:125-163. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

Jeanette I. Webster, Leonardo Tonelli, and Esther M. Sternberg§ National Institute of Mental Health, Section on Neuroimmune Immunology and Behavior, Bldg 36, Room 1A 23 (MSC 4020), 36 Convent Drive, Bethesda, Maryland 20892-4020; e-mail: [email protected]; [email protected]; [email protected]

Key Words glucocorticoid, immune response, inflammatory/autoimmune disease, HPA axis, cytokines ■ Abstract A reciprocal regulation exists between the central nervous and immune systems through which the CNS signals the immune system via hormonal and neuronal pathways and the immune system signals the CNS through cytokines. The primary hormonal pathway by which the CNS regulates the immune system is the hypothalamic-pituitary-adrenal axis, through the hormones of the neuroendocrine stress response. The sympathetic nervous system regulates the function of the immune system primarily via adrenergic neurotransmitters released through neuronal routes. Neuroendocrine regulation of immune function is essential for survival during stress or infection and to modulate immune responses in inflammatory disease. Glucocorticoids are the main effector end point of this neuroendocrine system and, through the glucocorticoid receptor, have multiple effects on immune cells and molecules. This review focuses on the regulation of the immune response via the neuroendocrine system. Particular details are presented on the effects of interruptions of this regulatory loop at multiple levels in predisposition and expression of immune diseases and on mechanisms of glucocorticoid effects on immune cells and molecules.

GENERAL INTRODUCTION The immune system has for many years been known to be influenced by glucocorticoids, and glucocorticoids have been used in the treatment of inflammatory diseases since the 1940s. In fact, Kendall, Reichstein, and Hench received the Nobel Prize for this discovery in 1950 (1). Since then extensive research has shown the pharmacological effects of glucocorticoids on many aspects of immune cell function (2, 3). However, until recently the fact that glucocorticoids play an ∗ The US Government has the right to retain a nonexclusive, royalty-free license in and to any copyright covering this paper. § Corresponding author.

125

5 Feb 2002

7:56

Annu. Rev. Immunol. 2002.20:125-163. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

126

AR

AR152-06.tex

AR152-06.SGM

LaTeX2e(2001/05/10)

P1: GJC

WEBSTER ET AL.

essential physiological role in regulation of the immune system in health and disease was not fully appreciated. An understanding of the physiological mechanisms involved in glucocorticoid secretion and their regulation of the immune system under both normal and disease conditions is fundamental to our understanding of the pathogenesis of inflammatory disease and ultimately for the development of effective therapies for such diseases. Recent studies have shown that this physiological regulation of the immune system by glucocorticoids is only one part of an extensive regulatory network between the central nervous system (CNS), neuroendocrine system, and immune system. This network of connections through nerve pathways, hormonal cascades, and cellular interactions allows the CNS to regulate the immune system locally at sites of inflammation, regionally in immune organs, and systemically though hormonal routes. In turn through similar connections, the immune system also regulates the CNS. During inflammation, cytokines produced at the inflammatory site can signal to the brain and produce the symptoms of sickness behavior and fever (4, 5). Cytokines are also expressed in areas of the brain, e.g., glia, neurons, and macrophages, and play a role in both neuronal cell death (6, 7) and survival (8). Cytokine-mediated neuronal cell death is thought to play an important role in several neuro-degenerative diseases, such as neuro-AIDS, Alzheimer’s, multiple sclerosis, stroke, and nerve trauma. In addition to this growth-factor role when expressed within the CNS, cytokines produced in the periphery can function as hormones and can stimulate the CNS by several mechanisms. They can pass the blood-brain barrier (BBB) at leaky points, for example at the organum vasculosum lamina terminalis (OVLT) or median eminence. They may be actively transported across the BBB in small amounts (9). In addition they can rapidly signal the CNS through the vagus nerve (10, 11). They can influence the brain by activation of second messengers, such as nitric oxide and prostaglandins, after binding to receptors on endothelial cells (12, 13). While a full understanding of this bidirectional communication between the CNS and immune systems is important, this chapter focuses only on the regulation of the immune system by the CNS, mainly though the neuroendocrine system and the adrenergic system. For further information on the regulation of the CNS by the immune system see the review by Mulla & Buckingham (14). The CNS regulates the immune system through two major mechanisms: (a) the hormonal stress response and the production of glucocorticoids, and (b) the autonomic nervous system with the release of noradrenalin. The CNS can also regulate the immune system locally via the peripheral nerves with release of neuropeptides such as substance P and locally produced corticotrophin-releasing hormone (CRH). This latter mechanism is not the focus of this review; for further information on these refer to the following articles (15, 16). The main regulator of the glucocorticoid effect on the immune system is the hypothalamic-pituitary-adrenal axis (HPA axis) (Figure 1). The main components of the HPA axis are the paraventricular nucleus (PVN) in the hypothalamus of the brain, the anterior pituitary gland located at the base of the brain, and the adrenal

5 Feb 2002

7:56

AR

AR152-06.tex

AR152-06.SGM

LaTeX2e(2001/05/10)

P1: GJC

127

Annu. Rev. Immunol. 2002.20:125-163. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

NEUROENDOCRINE REGULATION OF IMMUNITY

Figure 1 Diagram of the routes of communication between the brain and immune system, including the HPA axis, sympathetic nervous system, and cytokine feedback to the brain.

5 Feb 2002

7:56

Annu. Rev. Immunol. 2002.20:125-163. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

128

AR

AR152-06.tex

AR152-06.SGM

LaTeX2e(2001/05/10)

P1: GJC

WEBSTER ET AL.

glands. CRH is secreted from the PVN of the hypothalamus into the hypophyseal portal blood supply and stimulates the expression of adrenocorticotropin hormone (ACTH) in the anterior pituitary gland. ACTH then circulates in the bloodstream to the adrenal glands where it induces the expression and release of glucocorticoids. The HPA axis is subject to regulation both from within the central nervous system and from the periphery. Glucocorticoids, themselves, feed back to negatively regulate the HPA axis, exerting their negative feedback at the hypothalamic and pituitary level. The HPA axis can also be regulated by other factors such as the sympathetic nervous system, cytokines, and other neuropeptides, such as arginine vasopressin (AVP) (17). CRH is also negatively regulated by ACTH and itself, as well as by other neuropeptides and neurotransmitters in the brain, e.g., γ -aminobutyric acid-benzodiazopines (GABA-BDZ) and opioid peptide systems. CRH is positively regulated by serotonergic, cholinergic, and histaminergic systems (18). The connections between the neuroendocrine system and immune system provide a finely tuned regulatory system required for health. Disturbances at any level of the HPA axis or glucocorticoid action lead to an imbalance of this system and enhanced susceptibility to infection and inflammatory or autoimmune disease. Overstimulation of the HPA axis with excessive amounts of circulating glucocorticoids and overall suppression of immune responses lead to enhanced susceptibility to infection, whereas understimulation results in lower circulating levels of glucocorticoids and susceptibility to inflammation. Dysregulation may also occur at the molecular level and in this instance would result in glucocorticoid resistance at a molecular level leading to enhanced susceptibility to inflammation. A detailed understanding, therefore, by which the CNS and neuroendocrine systems regulate the immune system at the systemic, anatomical, cellular, and molecular levels will inform not only the pathogenesis and treatment of inflammatory/autoimmune conditions and infectious disease but also conditions predisposing to susceptibility and resistance to these illnesses.

EVIDENCE FOR ROLE OF ENDOGENOUS GLUCOCORTICOIDS IN REGULATION OF IMMUNE FUNCTION Changes in the levels of circulating glucocorticoids, such as during exercise and with circadian rhythms, are associated with changes in cytokine levels and production by leukocytes (19–23). Such studies provide circumstantial evidence that physiological fluctuations in glucocorticoid levels, in the range of those seen with circadian variations or stress, are associated with altered immune function. However, animal models have provided the strongest proof that endogenous glucocorticoids are essential physiological regulators of the immune response and inflammatory/autoimmune disease. Inbred rat strains, in which altered neuroendocrine responsiveness is associated with differential susceptibility and resistance to autoimmune/inflammatory

5 Feb 2002

7:56

AR

AR152-06.tex

AR152-06.SGM

LaTeX2e(2001/05/10)

P1: GJC

Annu. Rev. Immunol. 2002.20:125-163. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

NEUROENDOCRINE REGULATION OF IMMUNITY

129

disease, provide a naturalistic, genetically uniform system that can be systemically manipulated to test the role of neuroendocrine regulation of various aspects of immunity. Lewis (LEW/N) rats are an inbred strain of rats that are highly susceptible to development of a wide range of autoimmune/inflammatory diseases in response to a variety of antigenic or proinflammatory stimuli. Largely histocompatible Fischer (F344/N) rats are relatively resistant to these same illnesses after exposure to the same dose of antigens. These two strains also show related differences in HPA axis responsiveness, with inflammatory-susceptible LEW/N exhibiting a blunted response, compared to inflammatory-resistant F344/N rats with an excessive HPA response compared to outbred rats (24–28). Differences in the expression of hypothalamic CRH (26), pro-opiomelanocortin (POMC) (28), corticosterone-binding globulin (CBG) (27), and glucocorticoid receptor (GR) expression and activation (27, 29, 30) have been shown in these two rat strains. A variety of surgical or pharmacological HPA axis interventions in these strains of rats, as well as in mouse strains, alter the course and severity of inducible autoimmune/inflammatory disease. Thus, in F344/N rats, treatment with the glucocorticoid antagonist RU486 is associated with high mortality and development of arthritis in response to injection of streptococcal cell walls (25). Approximately 50% of rats die after infection with Salmonella typhimurium, but adrenalectomized rats suffer 100% lethality after infection with this bacteria (31). Similarly, in mice, adrenalectomy before infection with CMV virus results in lethality but glucocorticoid replacement prevents virus-induced lethality (32). The high rates of mortality within 12–24 h of exposure to this wide range of antigenic, proinflammatory, or infectious stimuli across species and strains in which the HPA axis has been interrupted pharmacologically or surgically, indicates the importance of an intact HPA axis to protect against septic shock. It is also possible in animal models to attenuate inflammatory disease by reconstituting the HPA axis, pharmacologically with glucocorticoids, or surgically by intracerebral fetal hypothalamic tissue transplantation. In LEW/N rats either treated with low-dose dexamethasone (25) or transplanted intracerebroventricularly with F344/N hypothalamic tissue (33), arthritis and carageenan inflammation are significantly attenuated. Experimental allergic encephalomyelitis (EAE) is inducible in LEW/N rats by immunization with myelin basic protein. After initial immunization these animals experience transient paralysis but then recover to some extent. During the development of the disease and recovery, the production of endogenous corticosterone increases, which is essential for survival of the animal. If this endogenous production is interrupted by adrenalectomy, the development of EAE is fatal. Whereas a subcutaneous steroid implant equivalent to the basal corticosterone levels does not reduce mortality, a dose equivalent to the EAE-induced corticosterone levels results in survival rates similar to that of normal animals, but EAE still develops. If a subcutaneous implant of even higher corticosterone levels is used, however, the animal will undergo complete remission (34). Such animal studies showing that interruption of the HPA axis predisposes to worse inflammation, while reconstitution attenuates autoimmune/inflammatory

5 Feb 2002

7:56

Annu. Rev. Immunol. 2002.20:125-163. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

130

AR

AR152-06.tex

AR152-06.SGM

LaTeX2e(2001/05/10)

P1: GJC

WEBSTER ET AL.

disease, lend support to the role of glucocorticoid imbalance in exacerbation of human autoimmune, inflammatory, allergic, or infectious diseases. Thus, a blunted HPA axis response to pituitary adrenal axis stimulation with CRH or insulin, or by psychological stress, has been shown in a variety of autoimmune diseases including rheumatoid arthritis, SLE, Sjogren’s syndrome, fibromyalgia, and chronic fatigue syndrome, as well as in allergic asthma and atopic dermatitis. Conversely, in humans, excess chronic stimulation of the hormonal stress response with concomitant chronically elevated glucocorticoids, as occurs in chronic stress situations is associated with an enhanced susceptibility to viral infection, prolonged wound healing, or decreased antibody production after vaccination (35–38). Such stress is experienced by caregivers of Alzheimer’s patients, students taking exams, couples during marital conflict, and Army Rangers undergoing extreme exercise and stress. One important mechanism by which activation of the HPA axis regulates these immune responses and severity of expression of resultant disease is through the effects of glucocorticoids at the molecular level though the glucocorticoid receptor (GR). The next sections describes this important receptor system and its regulation of target gene expression.

PHARMACOLOGICAL VERSUS PHYSIOLOGICAL EFFECTS OF GLUCOCORTICOID When considering the physiological relevance of the effects of glucocorticoids on immunity, it is important to recognize that glucocorticoids in pharmacological doses or forms exert different effects than they do under physiological conditions. Physiological concentrations of glucocorticoids in the range of 350 nmol/l to 950 nmol/l, such as occur during physical or psychological stress, result in modulation of transcription of genes involved in the inflammatory response, whereas pharmacological doses (higher concentration than physiological) result in a total suppression of the inflammatory response. There are also differences in potency of immune suppression by synthetic glucocorticoids, such as dexamethasone, versus the effects on immune response to natural glucocorticoids, such as hydrocortisone. For example, the synthetic glucocorticoid dexamethasone exerts a greater suppression of IL-12 than does the natural glucocorticoid hydrocortisone, which is consistent with the greater affinity of dexamethasone than hydrocortisone for GR (39).

MODULATION OF THE IMMUNE SYSTEM BY GLUCOCORTICOIDS There are two receptors for glucocorticoids, the glucocorticoid receptor and the mineralocorticoid receptor (MR). Corticosterone has a lower affinity for MR than for GR. Thus, at low levels, glucocorticoids bind preferentially to MR, and

5 Feb 2002

7:56

AR

AR152-06.tex

AR152-06.SGM

LaTeX2e(2001/05/10)

P1: GJC

Annu. Rev. Immunol. 2002.20:125-163. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

NEUROENDOCRINE REGULATION OF IMMUNITY

131

only at high levels, i.e., during stress, is GR occupied (39, 41). MR and GR, which are colocalized in some areas of the brain and in lymphocytes, can bind as heterodimers to DNA (42, 43), which could have implications for gene transcription or transrepression. In the brain, MR has been implicated in a proactive mode, in the prevention of disturbance of homeostasis, whereas GR has been suggested to work in a reactive mode in the recovery from a disturbance (39, 44). The primary receptor for glucocorticoids in immune cells is GR. The availability of glucocorticoids is also dependent on the expression of 11β-hydroxysteroid dehydrogenase, an enzyme responsible for the conversion of steroids from the active form, e.g., cortisol and corticosterone, into an 11-keto inactive form, e.g., cortisone and 11dehydrocortisone. Two forms of this enzyme exist. The type I enzyme is expressed in liver, brain, adipose tissue, lung, and other glucocorticoid-target tissues and catalyzes the regeneration of active glucocorticoids from the inactive 11-keto form. Conversely, the type II enzyme catalyzes the inactivation of glucocorticoids to the inert 11-keto form (45).

Glucocorticoid Receptor and Mechanism of Action The end point tissue effect of glucocorticoids is mediated by GR (NR3C1) (46). This is a member of the steroid and thyroid hormone receptor superfamily along with the progesterone, estrogen, mineralocorticoid, and thyroid receptors, and GR essentially is a ligand-dependent transcription factor. These receptors all have a similar structure that can be divided into three distinct regions (Figure 2). The N-terminal domain is involved in transactivation; the middle section is termed the DNA-binding domain (DBD) and is involved in DNA binding mediated via two zinc fingers; and the C-terminal domain or ligand-binding domain (LDB) is responsible for ligand binding as well as being also involved in transactivation, dimerization, and hsp90 binding (2, 47). Glucocorticoids circulate in the plasma associated with CBG or albumin. It is generally thought that glucocorticoids enter the cell by passive diffusion,

Figure 2 Structure of the glucocorticoid receptor.

5 Feb 2002

7:56

Annu. Rev. Immunol. 2002.20:125-163. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

132

AR

AR152-06.tex

AR152-06.SGM

LaTeX2e(2001/05/10)

P1: GJC

WEBSTER ET AL.

although there is evidence for an active transport out of the cell. GR is located in the cytoplasm in the unactivated state in a multiprotein complex containing hsp90 and immunophilins. These are thought to hold the GR in a conformation that is available to the ligand. Upon ligand binding, GR dissociates from this complex and translocates to the nucleus where it binds as a homodimer to target elements or glucocorticoid response elements (GREs) via its zinc fingers of the DBD. The bound GR homodimer can then modulate gene expression by modulating the basal transcription machinery either directly or via cofactors (Figure 3). GR has been involved in both the upregulation and downregulation of genes. Downregulation of gene expression can occur via so-called negative glucocorticoid response elements (nGRE), e.g., the POMC gene, but mostly gene repression by GR occurs via its interaction with other transcription factors such as AP-1 and NFκB (2, 48). It is noteworthy that the reverse can also occur, i.e., AP-1 and NFκB are able to repress GR function. When GR was first cloned, two clones for GR were found that differed in the C terminus (49). This second receptor, GRβ, is a splice variant of GR that is identical to GR but is lacking the last 50 amino acids. Instead it has a unique 15– amino acid C terminus. This receptor is located in the nucleus regardless of ligand

Figure 3 Schematic diagram of the mechanisms of action of the glucocorticoid receptor.

5 Feb 2002

7:56

AR

AR152-06.tex

AR152-06.SGM

LaTeX2e(2001/05/10)

P1: GJC

NEUROENDOCRINE REGULATION OF IMMUNITY

133

status, but it can also be found in the cytoplasm complexed to hsp90. It does not bind ligand and does not activate gene transcription but may act as a dominant negative receptor in vitro by forming transcriptionally inactive heterodimers with GRα (50–52). This mechanism of a dominant negative receptor is still under dispute as other studies have shown no effect of GRβ on GRα-mediated transactivation or transrepression (53–55).

Annu. Rev. Immunol. 2002.20:125-163. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

AP-1 GR can modulate gene expression via interaction with other factors, such as NFκB and AP-1. AP-1 consists of a heterodimer of the oncoproteins c-Fos and c-Jun that bind to an AP-1 consensus binding site in DNA to activate gene transcription in response to ligands such as phorbol esters. Although the high-affinity Fos/Jun heterodimers predominate, the lower-affinity Jun/Jun homodimers are also capable of binding to the AP-1 site. Glucocorticoids can reduce the DNA-binding ability of the Fos and Jun complexes by protein-protein interactions and thereby repress AP-1-mediated gene activation (56, 57) (Figure 4). This can also occur through a composite GRE. This sequence contains a binding site for the receptor and also for a nonreceptor factor. For example, GR was shown to repress AP-1 activity at the composite GRE in the promoter of the plfG gene; this repression was mediated by the N-terminal domain of GR and could not be mediated by MR (58, 59). One example of a gene where a composite HRE is involved in the repression of the gene by glucocorticoids is the hypothalamic hormone CRH. The promoter of this gene contains an AP-1 site closely located to a GRE, and repression of the AP-1-activated gene transcription by glucocorticoids has been shown (60). This could be the mechanism by which glucocorticoids exert their negative feedback on the hypothalamus to downregulate the HPA axis.

NFκB Glucocorticoids play a role in immunosuppression through their repression of NFκB, which is a major factor involved in the regulation of cytokines and other immune responses. [For comprehensive reviews on NFκB, see Baldwin (61), Ghosh et al. (62), and McKay & Cidlowski (48)]. NFκB was originally described as a dimer of two proteins p65 (RelA) and p50 (NF-κB1), but a whole family of these proteins has now been described. These proteins all contain a highly conserved 300–amino acid domain termed the Rel homology domain (RHD), which is important for DNA binding and nuclear translocation and dimerization. The proteins p65 (Rel A), C Rel, and Rel B contain a C-terminal transactivation domain, whereas p50 (NF-κB1) and p52 (NF-κB2) are present as precursors (p105 and p100, respectively), which contain a C-terminal IκB-like region that is removed by proteolysis. The ability of these proteins to form various dimer partners and the specificity of these dimers for different DNA sequences may be fundamental to the cellspecific responses mediated by NFκB. These proteins are present in the cytoplasm as complexes together with the inhibitory protein IκB. This is also now known to be part of a family of proteins that include IκBα, IκBβ, IκBγ , IκB-R, and the

5 Feb 2002

7:56

Annu. Rev. Immunol. 2002.20:125-163. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

134

AR

AR152-06.tex

AR152-06.SGM

LaTeX2e(2001/05/10)

P1: GJC

WEBSTER ET AL.

Figure 4 Schematic diagram of the mechanism(s) by which the glucocorticoid receptor inhibits the action of NFκB and AP-1.

C-termini of p105/p50 and p100/p52. These all contain multiple ankyrin repeats, which are involved in protein-protein interactions between the IκB and NFκB, and a C-terminal PEST sequence, which is a signal for protein degradation (48). NFκB is normally located in the cytoplasm associated with the inhibitor IκB protein. Upon activation, IκB is phosphorylated, ubiquinated, and then degraded. The free NFκB can then translocate to the nucleus where it activates gene expression. NFκB can be stimulated by a variety of factors, including proinflammatory cytokines such as TNF-α and IL-1, physical or oxidative stress, and bacterial or viral proteins (48, 63) (Figure 4). There have been two schools of thought regarding GR-mediated repression of NFκB, and data exist to support both. One is that glucocorticoids via GR induce the expression of the inhibitory protein IκB that then sequesters NFκB in the cytoplasm and prevents it from translocating to the nucleus and inducing gene

5 Feb 2002

7:56

AR

AR152-06.tex

AR152-06.SGM

LaTeX2e(2001/05/10)

P1: GJC

Annu. Rev. Immunol. 2002.20:125-163. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

NEUROENDOCRINE REGULATION OF IMMUNITY

135

activation (63, 64). This mechanism of repression of NFκB by GR may be limited to certain cell types, particularly monocytes and lymphocytes. The other model suggests that there is a physical interaction or cross-talk between NFκB and GR that prevents gene expression (63, 65–71). In some cases GR-induced transcription is not required, as the anti-glucocorticoid RU486 is also capable of inhibiting NFκB to some extent (67, 72). The region of the GR required for GR repression of NFκB has been located to the zinc finger region of the DNA-binding region of GR (63, 67). This region has been further defined as two amino acids in the C-terminal zinc finger that are absolutely critical for GR-mediated repression of NFκB; these residues are not involved in GR-mediated repression of AP-1 (73). It is plausible that these two models are not mutually exclusive or that they are dependent on cell type. One experiment in the A549 pulmonary epithelial cell line has shown that both these models exist in the same system. In this case, dexamethasone does induce IκB expression, but when protein synthesis and therefore IκB production is blocked, GR is still able to repress NFκB-mediated gene activation although not to the same extent as when protein synthesis is allowed (74). It has also been suggested that glucocorticoid repression of NFκB activity could be caused by competition between GR and NFκB for limited cofactors such as CREB-binding protein (CBP) and steroid receptor coactivator 1 (SRC-1), as it has been shown in Cos cells that this repression can be alleviated by excess cofactor (75). McKay & Cidlowski, however, showed that CBP did not mediate GR repression of NFκB by a competitive model, but rather that CBP functioned as an integrator to enhance the physical interaction between GR and NFκB (76). The catalytic subunit of protein kinase A (PKAc) has also been suggested to promote the cross-talk between GR and NFκB (77).

REGULATION OF IMMUNE-RELATED GENES Glucocorticoids regulate a wide variety of immune cell functions and expression of immune molecules through the molecular mechanisms described above. Thus, glucocorticoids modulate cytokine expression, adhesion molecule expression and immune cell trafficking, immune cell maturation and differentiation, expression of chemoattractants and cell migration, and production of inflammatory mediators and other inflammatory molecules [for reviews see Barnes (78) and Adcock (2)].

Cytokine Expression Glucocorticoids modulate the transcription of many cytokines. They suppress the proinflammatory cytokines IL-1, IL-2, IL-6, IL-8, IL-11, IL-12, TNF-α, IFN-γ , and GM-CSF while upregulating the anti-inflammatory cytokines IL-4 and IL-10. PROINFLAMMATORY CYTOKINES Glucocorticoids mediate the anti-inflammatory response by downregulating the expression of proinflammatory cytokines such

5 Feb 2002

7:56

Annu. Rev. Immunol. 2002.20:125-163. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

136

AR

AR152-06.tex

AR152-06.SGM

LaTeX2e(2001/05/10)

P1: GJC

WEBSTER ET AL.

as IL-1, IL-6, and TNF-α. Glucocorticoids have been shown to downregulate the transcription of IL-1β, to destabilize IL-1β mRNA (79–81), and to downregulate IL-1α (82). They also upregulate the expression of the decoy IL-1 receptor, IL-1R II (83). Glucocorticoids regulate IL-6 production not only directly, but also indirectly, through their effects on the cytokines that regulate IL-6. IL-1 or TNF induction of IL-6 is mediated via NFκB and by another nuclear factor, NF-IL6, in transfection studies of these nuclear factors and the IL-6 promoter in HeLa and F9 cells. This IL-1-induced induction of IL-6 can be inhibited by glucocorticoid-activated GR via protein-protein interactions at the C-terminal transactivation domain of NFκB (65, 69, 84). This repression does not require protein synthesis and does not affect the DNA-binding capacity of NFκB (69). Glucocorticoids also inhibit expression of IL-11, a member of the IL-6 cytokine family, by inhibition of gene expression and by destabilization of mRNA (85). Evidence for these molecular mechanisms of GC regulation of cytokines has appeared in vivo in disease states. In rheumatoid arthritis patients, administration of glucocorticoids leads to a decrease in the release of TNF-α into the bloodstream (86). The LPS-stimulated production of TNF-α in monocytes is reduced by glucocorticoid treatment. The promoter of TNF-α does not contain a typical GRE site, and so this repression by glucocorticoids may occur via other factors involved in the regulation of TNF-α which have binding sites in the promoter, such as NFκB and AP-1 (87). However, some evidence suggests that the glucocorticoid-mediated repression of TNF-α could occur at the level of translation rather than transcription (88). The transcription of other proinflammatory cytokines is also affected by glucocorticoids. IFN-γ expression in spleen cells, and blood monocytes is inhibited by glucocorticoids (89, 90). Dexamethasone inhibits basal IL-8 in airway epithelial cells by destabilizing the mRNA via new protein synthesis (91). The expression of the Th1 cytokine IL-12 and its receptor is downregulated by glucocorticoids (92–94). Effects of glucocorticoids on IL-12 are fully discussed in the section on Th1-Th2 shift. IL-2 expression in spleen cells is inhibited by glucocorticoids (89), and dexamethasone inhibits the ionomycin- and TPA-induced IL-2 expression in FJ8.1 cells. With the promoter for IL-2, the dexamethasone-induced repression inhibits the binding of NFκB and reduces the binding of AP-1 to DNA. Glucocorticoids not only affect the action of cytokines by affecting their expression, but they also can inhibit their signaling mechanisms. Thus IL-2 signaling via the Jak-STAT cascade can be inhibited by dexamethasone (95). Granulocyte-macrophage colony stimulating factor (GM-CSF) is a proinflammatory mediator. IL-1β-stimulated GM-CSF in bronchial epithelial cells is inhibited by dexamethasone by transcriptional mechanisms (96). ANTIINFLAMMATORY CYTOKINES The major Th2 cytokine is IL-10. In blood monocytes varying effects of glucocorticoids, both synthetic and natural, on IL-10 have

5 Feb 2002

7:56

AR

AR152-06.tex

AR152-06.SGM

LaTeX2e(2001/05/10)

P1: GJC

Annu. Rev. Immunol. 2002.20:125-163. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

NEUROENDOCRINE REGULATION OF IMMUNITY

137

been reported that are dependent on the dose. Low physiological doses of glucocorticoids suppress IL-10 expression, whereas high pharmacological doses enhance IL-10 expression (39, 97a). IL-4 is a Th2-associated cytokine that is important for the proliferation of mast cells and attraction of eosinophils to areas of inflammation during response to IgE stimuli or allergic reactions. Dexamethasone downregulates IL-4 mRNA in mast cells (98, 99) and in lymphocytes (82). The effect of dexamethasone on IL-4 production in T cells is a matter of some controversy and can be at least partially explained by consideration of the doses used experimentally. As IL-4 is a Th2-associated cytokine and since glucocorticoids induce a shift from Th1Th2 immunity (see later section), one might expect dexamethasone to induce IL-4 expression. Indeed, in the lymph nodes and spleen of mice IL-4 is induced in response to physiological concentrations of glucocorticoids both in vivo and in vitro (100, 101). On the other hand, consistent with the efficiency of glucocorticoids in the treatment of allergic diseases, data also exist showing that stress levels of dexamethasone downregulate IL-4 expression in T cells (89).

Cell Adhesion Molecules and Immune Cell Trafficking Glucocorticoids reduce the trafficking of leukocytes to areas of inflammation. This occurs via the downregulation of protein molecules involved in the attraction and adhesion of leukocytes to these areas. Dexamethasone inhibits the expression of intracellular adhesion molecule 1 (ICAM-1), endothelial-leukocyte adhesion molecule 1 (ELAM-1) (102), and vascular adhesion molecule 1 (VCAM-1) (103). Glucocorticoids can inhibit the expression of ICAM-1 through inhibition of the NFκB site in the promoter of the ICAM-1 gene (66, 71, 74). Glucocorticoids inhibit TNF-α–induced VCAM-1 expression but not ICAM-1 in bronchial epithelial cells. NFκB sites have been identified in the promoter of VCAM-1, and it has been suggested that glucocorticoid repression of this gene may occur through GR-mediated repression of NFκB (103). E-selectin is an endothelial cell surface adhesion molecule that is important for the recruitment of leukocytes from the blood. It is upregulated by IL-1β and TNF-α and its expression can be repressed by glucocorticoids. Repression by glucocorticoids does not affect NFκB translocation or binding to DNA, which suggests interference in transcriptional activation of NFκB (70). L-selectin is involved in the attachment of blood leukocytes during inflammation, and its expression is decreased by glucocorticoid treatment in bone marrow cells and polymorphonuclear cells (104).

Chemoattractants and Cell Migration Cytokine-induced neutrophil chemoattractant (CINC)/gro is a chemoattractant for neutrophils and is therefore important for the accumulation of neutrophils at sites of inflammation. It is also proposed to be a member of the family that includes IL-8. mRNA for CINC/gro is induced by IL-1β via NFκB, and this induction can

5 Feb 2002

7:56

Annu. Rev. Immunol. 2002.20:125-163. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

138

AR

AR152-06.tex

AR152-06.SGM

LaTeX2e(2001/05/10)

P1: GJC

WEBSTER ET AL.

be inhibited by glucocorticoids, which prevent the translocation of NFκB to the nucleus and therefore prevent DNA binding and gene transcription (105). IL-5 is a chemoattractant for eosinophils and is important for the accumulation of eosinophils in inflammatory diseases such as asthma. IL-5 mRNA can be downregulated by glucocorticoids in mast cells (99) and T cells (106). IL-5 can be induced by TCR and IL-2 by different mechanisms, and both of these can be repressed by glucocorticoids. The mechanism of glucocorticoid repression of TCR-induced IL-5 possibly occurs through AP-1 and NFκB, but these factors are not involved in the glucocorticoid repression of IL-2-induced IL-5 (107). RANTES (regulated upon activation normal T cell expressed and secreted) is a chemoattractant involved in the recruitment of eosinophils to areas of inflammation. Dexamethasone inhibits mRNA expression in activated and nonactivated cells (108, 109). The cytokines monocyte chemoattractant protein 1 (MCP-1), MCP2, and MCP-3 are downregulated by dexamethasone (108). Downregulation of MCP-1 occurs posttranscriptionally by destabilization of the mRNA at the 50 end, although the exact mechanism is unknown (110). Eotaxin is an eosinophil chemoattractant involved in the accumulation of eosinophils in the airways in response to allergic stimuli. Eotaxin mRNA is stimulated by cytokines such as TNF-α and IL-1β and this cytokine-induced induction is repressed by glucocorticoids (111).

Production of Inflammatory Mediators Glucocorticoids also affect the production of inflammatory mediators including prostaglandins and nitric oxide. Glucocorticoids suppress prostaglandin synthesis at sites of inflammation by several mechanisms. The key stages during the synthesis of prostaglandin are the release of arachadonic acid from membrane phospholipids, which is catalyzed by phospholipase A2 (PLA2), and the subsequent conversion to PGH2 by the action of cyclooxygenase (COX). There are two isoforms of PLA2, secretory PLA2 and cytosolic PLA2. There are also two COX enzymes; COX-1 is constitutively expressed, and COX-2 is inducible and thought to be involved in stimuli-induced prostaglandin synthesis. COX-2 induction by inflammatory stimuli is mediated though NF-IL6 and NFκB regulatory sequences in its promoter (112). Glucocorticoids exert their effect by repression of cytosolic PLA2 and COX2 mRNA levels (113–115). Nitric oxide synthase II (NOS II) is the enzyme that produces nitric oxide (NO). NO has been implicated in autoimmune and inflammatory responses. The promoter of NOS II contains a binding site for NFκB, which is required for the cytokine induction of this gene. Glucocorticoids repress the cytokine induction of NOS II, and this repression does not correspond with an induction of the IκB gene, thereby implicating a method of protein-protein interaction for this repression (68). iNOS is an inducible form of nitric oxide synthase and is another enzyme involved in the synthesis of nitric oxide. Dexamethasone inhibits transcription of the iNOS gene in rat hepatocytes by the induction of IκB (116).

5 Feb 2002

7:56

AR

AR152-06.tex

AR152-06.SGM

LaTeX2e(2001/05/10)

P1: GJC

Annu. Rev. Immunol. 2002.20:125-163. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

NEUROENDOCRINE REGULATION OF IMMUNITY

139

Lipocortin 1 (or annexin 1) is expressed in many immune cells and is also expressed in the neuroendocrine system in the hypothalamus and pituitary. It has anti-inflammatory actions although these mechanisms are not fully understood. One way in which lipocortin 1 exerts its anti-inflammatory action in the periphery is by the inhibition of prostaglandin synthesis via the inhibition of release of arachadonic acid, although the exact mechanism is not known. Glucocorticoids have a dual action on lipocortin 1. Lipocortin 1 is normally expressed and localized mainly in the cytoplasm, but it is also localized to the external surface of cell membranes. Upon glucocorticoid exposure, the amount of lipocortin 1 on the external membrane rapidly increases. This externalization is followed by an increase in lipocortin 1 synthesis by the cells in response to glucocorticoids (117).

Other Factors Involved in the Inflammatory Response The β 2-adrenoreceptor is involved in the adrenergic control of the immune system (see later section). Glucocorticoids increase the transcription of the β 2-adrenoreceptor in many cell types, including airway epithelial cells. This transcriptional regulation occurs via a GRE sequence in the promoter of the gene (118, 119). Other receptors involved in the regulation of the immune system are regulated by glucocorticoids. The transcription of the NK1-receptor, the receptor for substance P, is inhibited by glucocorticoids probably through an AP-1 site (120). The NK2receptor is also downregulated by glucocorticoids (121).

EFFECTS ON IMMUNE CELLS Glucocorticoids, in general, suppress maturation, differentiation, and proliferation of immune cells involved in all aspects of immunity, including innate, T cell, and B cell function and chronic allergic reactions.

Innate Immunity Monocytes are produced in the bone marrow. Once mature they leave, circulate in the bloodstream, and after a time enter the tissues and become macrophages. During inflammation, circulating monocytes migrate to the inflammatory site where they become activated. Glucocorticoids interfere with the protective and defense mechanisms of activated macrophages, rather than resident macrophages. This occurs via their effects on cytokines and other inflammatory mediators such as prostaglandin synthesis (122). Glucocorticoids at pharmacological concentrations induce apoptosis in macrophages and monocytes; and the pro-inflammatory cytokine IL-1β may be involved in prevention of apoptosis, whereas the antiinflammatory cytokines IL-4 and IL-10 induce apoptosis in these cells (81). Thus, pharmacological or stress levels of glucocorticoids reduce circulating numbers of monocytes, inhibit secretion of IL-1, IL-6, TNF-α, and monocyte chemotactic activating factors, and impair synthesis of collagenase, elastase, and tissue plasminogen activator (123).

5 Feb 2002

7:56

Annu. Rev. Immunol. 2002.20:125-163. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

140

AR

AR152-06.tex

AR152-06.SGM

LaTeX2e(2001/05/10)

P1: GJC

WEBSTER ET AL.

Neutrophils are involved in the tissue damage that occurs during the innate inflammatory phase of inflammatory disease. In response to injury, neutrophils migrate out of the circulating bloodstream, extravasate between endothelial cells, and move to the site of inflammation. Glucocorticoids affect the activation of neutrophils and also the functions of neutrophils, such as chemotaxis, adhesion, transmigration, apoptosis, and phagocytosis (123, 124). Glucocorticoid regulation of these functions in neutrophils is due to numerous components including regulation of cytokines. However, lipocortin 1 also plays a pivotal role in glucocorticoidinduced responses. Glucocorticoids have a dual effect on neutrophils. On one hand pharmacological doses are inhibitory and suppress the inflammatory response caused by neutrophil activation and migration. On the other hand, neutrophils are required for the response to bacterial infections, and as such their circulating numbers are increased by pharmacological doses of glucocorticoids through inhibition of apoptosis (124).

Th1-Th2 Glucocorticoids induce a shift from a Th1 to a Th2 pattern of immunity. Th1 immunity or cellular immunity is characterized by the expression of proinflammatory cytokines, such as IFN-γ , IL-2, and TNF-β, which lead to the differentiation of macrophages, natural killer cells (NK cells), and cytotoxic T cells that are involved in phagocytosis and destruction of invading bacteria or foreign bodies. Th2 immunity or humoral immunity is characterized by the production of antiinflammatory cytokines such as IL-4, IL-10, and IL-13, resulting in the differentiation of eosinophils, mast cells, and B cells, which lead to an antibody-mediated defense against foreign antigens. The main inducer of Th1 immunity is IL-12. This can induce the expression of IFN-γ and inhibit the expression of IL-4, which plays a critical role in the immune response. [For a detailed review on IL-4, see Nelms et al. (125)]. As previously mentioned, glucocorticoids can affect the transcription of many cytokines, generally upregulating antiinflammatory cytokines and downregulating proinflammatory cytokines, thereby causing a shift from Th1 to Th2 immunity. The glucocorticoid-induced shift of Th1 to Th2 may be due mainly to downregulation of the Th1 cytokines, thus allowing dominant expression of the Th2 cytokines (126, 127). They modulate the expression of IL-12 or its receptor, and this regulation of IL-12 is thought to be a major mechanism by which glucocorticoids mediate the Th1-Th2 shift (94). This glucocorticoid-mediated downregulation of IL-12 occurs by inhibition of Stat4 phosphorylation in the signaling cascade downstream of IL-12, and this inhibition is mediated by GR. This is specific for the Th1-mediated immunity because phosphorylation of Stat6, which is involved in the signal cascade from IL-4, is not affected by glucocorticoids (93). Glucocorticoids downregulate the IL-12 receptor β1- and β2-chains, but this downregulation occurs after 3 days of treatment, whereas the inhibition of

5 Feb 2002

7:56

AR

AR152-06.tex

AR152-06.SGM

LaTeX2e(2001/05/10)

P1: GJC

Annu. Rev. Immunol. 2002.20:125-163. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

NEUROENDOCRINE REGULATION OF IMMUNITY

141

Stat4 phosphorylation is seen within 2 h, suggesting that IL-12 receptor downregulation occurs in response to prolonged glucocorticoid exposure (92, 93). Several other neural factors besides glucocorticoids can also induce a Th1-Th2 shift; these neural factors include the sympathetic neuropeptides, noradrenalin, and adrenalin (94). Glucocorticoids differentially affect the survival of Th1 and Th2 cells. Th2 NK1.1+ T cells, which produce IL-4, are resistant to dexamethasone-induced apoptosis, and this resistance may be due to the expression of the proto-oncogene Bcl-2 (128). Alterations in Th1-Th2 immunity are characteristic of some autoimmune diseases. Rheumatoid arthritis, multiple sclerosis (MS), and type I diabetes mellitus are examples of autoimmune diseases in which there is a shift toward Th1-mediated immunity with an excess of IL-12 and TNF-α production. Systemic lupus erythematosus (SLE) is shifted toward Th2-mediated immunity with an excess of IL-10 production (94). In situations where there is an excess of glucocorticoid production, e.g., in animal models with a hyperactive HPA axis (F344/N rats) or in women in the third trimester of pregnancy, there is a relative resistance to Th1-associated autoimmune diseases. Conversely, animals with a lack of glucocorticoids or a hypoactive HPA axis (LEW/N rats) are susceptible to Th1-associated autoimmune diseases (94, 129). LEW/N rats are susceptible to EAE, a Th1-mediated autoimmune disease. Spontaneous recovery of these animals is correlated with an increase in glucocorticoids, whereas adrenalectomy results in fatal regression of the disease but can be reversed by administration of glucocorticoids, indicating the shift from Th1 to Th2 immunity (34).

Allergy Eosinophils are critical in the response to allergic stimuli, and glucocorticoids both systemically and topically applied are effective in the treatment of these diseases. Pharmacological doses of glucocorticoids reduce the circulating numbers of eosinophils probably through many factors including IL-5, IL-3, and GM-CSF. Glucocorticoids also sequester eosinophils in primary and secondary lymphoid tissues, induce apoptosis, and inhibit the recruitment of eosinophils to areas of inflammation (123, 130, 131). Basophils are the least abundant circulating leukocyte but are important in IgEmediated allergic inflammation. Their activation leads to the synthesis and release of lipid mediators such as leukotriene C4 (LTC4), which induces changes in muscle and vascular tissue. Glucocorticoids reduce circulating numbers of basophils, impair histamine and leukotriene release, and inhibit basophil migration (123, 131). Mast cells are important for the mediation of inflammatory disease particularly in response to IgE. Glucocorticoid treatment results in a decrease in mast cells in airways, but the mechanism for this is not certain, and a direct inhibitory effect on mediator release has been disproven (132, 133).

5 Feb 2002

7:56

142

AR

AR152-06.tex

AR152-06.SGM

LaTeX2e(2001/05/10)

P1: GJC

WEBSTER ET AL.

Annu. Rev. Immunol. 2002.20:125-163. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

Resident Tissue Immune Cells Dendritic cells are involved in antigen presentation, and this could assist in the Th1-Th2 shift. Glucocorticoids reduce the number of dendritic cells in an animal model and also in human nasal mucosa after topical glucocorticoid application (134, 135). Epithelial cells are involved in inflammation in asthmatic airways, and they are also a target for glucocorticoid therapy in this disease. As described earlier, glucocorticoids inhibit the expression of cytokines, chemoattractants, and mediators of inflammation in the epithelial cells of the airways. Glucocorticoids reduce the inflammatory mediator–induced leakage of plasma proteins into the alveolar bronchi (136). Interestingly, endothelial cells of the airways show the highest expression of GR in human lung (137). Mucosal cells secrete mucus during inflammation. Glucocorticoids act directly on these cells and inhibit the expression of MUC2 and MUC5AC, the genes that encode mucin (138, 139). Glucocorticoids also inhibit mucus secretion via their inhibition of inflammatory mediators.

Effects of Glucocorticoids on Lymphocyte Selection at the Cellular and Immune Organ Level Glucocorticoids play an important role in lymphocyte selection via several mechanisms, including apoptosis and thymocyte differentiation during development. Apoptosis is a mechanism of programmed cell death, which is necessary for a homeostasis to exist between cell death and proliferation. For a review on apoptosis, see King & Cidlowski (140). Glucocorticoids induce G1 arrest and apoptosis in lymphoid cells. There appear to be two mechanisms for glucocorticoidmediated apoptosis in thymocytes, one for proliferating thymocytes and one for nonproliferating thymocytes. Although separate mechanisms, these two pathways do share some common features such as the activation of GR and the activation of caspases. The mechanism of glucocorticoid-mediated cell cycle arrest and apoptosis is not fully understood, but some cell cycle genes, e.g., c-myc, cyclin D3, and Cdk4, are regulated by glucocorticoids (140, 141). Glucocorticoids upregulate the expression of the cyclin-dependent kinase inhibitor, p57Kip2, which is important in this glucocorticoids-mediated cell cycle arrest (142). In T cells, glucocorticoid-induced apoptosis is dependent on protein synthesis, but there is also evidence that it is dependent on repression of genes, such as c-myc. Comparison of the glucocorticoid-sensitive T-cell line 6TG1.1 and the glucocorticoid-resistant variant ICR27TK.3, which contains GR that is incapable of gene activation, indicates that repression of AP-1-induced gene activity is not involved in glucocorticoid-induced apoptosis in these cells. DNA fragmentation in the sensitive 6TG1.1 cells was inhibited by cyclohexamide, indicating that protein synthesis is required for glucocorticoid-induced apoptosis. Because inhibitory IκB molecule is induced by dexamethasone in these cells and because c-myc is regulated by NFκB, it has been suggested that glucocorticoid-induced apoptosis could occur APOPTOSIS

5 Feb 2002

7:56

AR

AR152-06.tex

AR152-06.SGM

LaTeX2e(2001/05/10)

P1: GJC

NEUROENDOCRINE REGULATION OF IMMUNITY

143

Annu. Rev. Immunol. 2002.20:125-163. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

by repression of gene transcription of the c-myc gene via glucocorticoid-induced transcription of IκB or another similar inhibitory factor (64). Glucocorticoid-mediated apoptosis has also been described in monocytes. This occurs via the death receptor, CD95, a cascade for apoptosis that does not occur in T cells or thymocytes in response to glucocorticoids. In monocytes, glucocorticoids increase the expression of membrane CD95 and the ligand, CD95L, and also enhance their release. This activates a cascade involving caspase 8 and caspase 3 to induce apoptosis (143). EFFECTS OF GLUCOCORTICOIDS ON THYMOCYTE DIFFERENTIATION AND SELECTION

During the process of thymocyte differentiation, immature CD4− CD8− thymocytes rearrange the β gene locus of the T cell receptor (TCR) to express pre-TCR. These then divide and begin to express CD4+ and CD8+ and rearrange the α gene of TCR to express mature α/β TCR but at low levels. These CD4+ CD8+ low TCR cells then undergo selection, either positive or negative selection. Precisely what factors determine positive versus negative selection is not understood. In vitro in the presence of glucocorticoids alone, thymocytes/hybridomas undergo glucocorticoid-mediated apoptosis, whereas in the absence of glucocorticoids and the presence of activated TCR, these cells undergo activated-mediated apoptosis. Culture in the presence of both results in survival, and this balance between glucocorticoids and TCR has been termed mutual antagonism. To explain how selection occurs, it has been proposed that where the CD4+ CD8+ low TCR thymocytes do not recognize self-antigen/major histocompatability complex (MHC), there is a greater ratio of GR to TCR stimulation, and these cells will die of neglect through glucocorticoid-mediated apoptosis. If these thymocytes strongly recognize self-MHC, with a resultant higher ratio of TCR to GR activation, then these cells will die by negative selection through activated-mediated apoptosis. If there is an intermediate recognition of the MHC, a balance exists between the GR and TCR activation, and these cells will be selected by positive selection and will further differentiate into mature CD4+ or CD8+ thymocytes (144, 145). Evidence for this mutual antagonism has been provided by studies in which inhibiting glucocorticoids results in thymocytes with a low-to-moderate avidity for self-MHC being forced into activated-mediated apoptosis (negative selection), rather than being rescued and undergoing positive selection as would occur in the presence of glucocorticoids (146). Furthermore, in these conditions thymocytes that do not recognize self-MHC are rescued rather than undergoing glucocorticoid-mediated apoptosis (147). Thymocyte selection is active during fetal and neonatal period, but the levels of glucocorticoids circulating in the fetus are low because some level of protection from maternal glucocorticoids is provided by the placenta that contains high levels of the GC-catabolizing enzyme 11β-hydroxysteroid dehydrogenase. This apparent lack of glucocorticoids in the fetus and the necessity of them for thymocyte differentiation has led to the hypothesis that there may be local production of glucocorticoids in the thymus during development (145). Indeed, cultured

5 Feb 2002

7:56

Annu. Rev. Immunol. 2002.20:125-163. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

144

AR

AR152-06.tex

AR152-06.SGM

LaTeX2e(2001/05/10)

P1: GJC

WEBSTER ET AL.

thymic epithelium cells do produce pregnenolone and deoxycorticosterone, and this production can be increased by ACTH. This production of glucocorticoids was specific to thymic epithelium cells and was not observed in cultures of thymocytes, macrophages, or dendritic cells. The P450 hydroxylase enzymes involved in the production of glucocorticoids, specifically corticosterone from cholesterol, have been shown in thymic epithelium cells (144, 145, 148). Murine thymus also contains all the enzymes required for the production of glucocorticoids from cholesterol (149). In hybridomas, activated-mediated apoptosis is caused by the upregulation of the ligand for Fas, Fas ligand (FasL). Glucocorticoids prevent this upregulation of FasL, probably by the glucocorticoid-induced leucine zipper (GILZ) gene, but do not interfere with Fas itself. Another mechanism by which glucocorticoids may inhibit activated-mediated apoptosis is the activation of glucocorticoidinduced TNFR family related (GITR), which can inhibit CD3-mediated but not Fas-mediated apoptosis. In some diseases there could be a shift in this balance. Thus, in cases where glucocorticoids are overexpressed, this balance could be shifted such that a higher level of activated TCR cells survive and thus a greater population of thymocytes that recognize self could survive (145). Indeed, in autoimmune disease models in animals and in patients with the autoimmune disease multiple sclerosis, there is a higher basal level of glucocorticoids (145, 150). The involvement of glucocorticoids in thymocyte differentiation and selection is still a matter of some controversy. Generation of transgenic antisense mice for the 30 untranslated region of GR under the control of a T cell–specific promoter led to mice with a twofold reduction of GR mRNA and protein. The thymus of transgenic homozygotes were 90% smaller than controls, and there was a decrease in CD4+ CD8+ thymocytes and a secondary decrease in CD4+ CD8− and CD4− CD8+ thymocytes, which indicates the necessity of glucocorticoids in thymocyte development and differentiation (144, 145). However, in similar transgenics under the control of a different promoter (a neurofilament promoter), GR levels were decreased in all tissues, and there was an associated increase in circulating ACTH and corticosterone. Unlike the T cell–specific antisense animals, no difference in CD4+ CD8+ thymocytes was found (151). In the GR knockin mice, which are deficient in dimerization and are therefore unable to activate gene transcription by GR, the thymocyte population was described as being normal although glucocorticoid-mediated apoptosis was reduced and no extensive analysis was performed (152). Recently, GR knockout mice have been described that die at birth [day 19 of embryogenesis (E19)]. However, during embryogenesis, i.e., up to E18, these animals have apparently normal thymocyte populations and T cell development, suggesting that GR in this case in not necessary for the entire process of T cell development and differentiation (153, 154). The reason for these differences between animal models is not completely clear, but one significant difference between these different transgenic animals and knockin animals may result from secondary effects of the disrupted HPA axis (145).

5 Feb 2002

7:56

AR

AR152-06.tex

AR152-06.SGM

LaTeX2e(2001/05/10)

P1: GJC

NEUROENDOCRINE REGULATION OF IMMUNITY

145

DISRUPTION OF THE HPA AXIS AND DISEASE

Annu. Rev. Immunol. 2002.20:125-163. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

Animal Models Animal models have been useful in understanding the pathogenesis of autoimmune/inflammatory disease. Many animal models exist in which a blunted HPA axis response has been associated with susceptibility to autoimmune/inflammatory diseases. These include the obese strain (OS) chicken as a model for autoimmune thyroiditis (155), certain mouse lupus (SLE) models (156, 157), and the inbred rat strains, LEW/N and F344/N rats, as described earlier. For a review on these and other animal models for inflammatory diseases, refer to Jafarian-Tehrani & Sternberg (158). It should be noted, however, that many genes and many chromosome regions, each with small effect, contribute to susceptibility to autoimmune diseases, such as SLE (158a) and inflammatory arthritis (158b). While some of these linkage regions contain genes that regulate the HPA axis (158c), many are unrelated or unknown. Such genetic linkage and segregation studies also indicate that environmental factors play a large role in variance of disease expression. Thus, many different genetic and environmental factors contribute to autoimmune susceptibility in addition to HPA or neuroendocrine factors.

Human Diseases HPA axis responsiveness differs greatly among individuals. Even in healthy individuals considerable intra-individual variability appears in HPA axis responses. Normal healthy volunteers have been subgrouped into high or low responders depending on the response of their HPA axis to stimuli (159, 160). Disruptions in the HPA axis or glucocorticoid response could occur at many levels, including at the levels of the hypothalamus, pituitary, or adrenals, with changes in the expression of CRH, ACTH or cortisol, or changes in the sensitivity of the system to stimuli or suppressive factors. Once cortisol reaches its target tissue, there are then many steps to inducing gene activation, including entry into the cell, binding to GR, dimerization, translocation to the nucleus, and interaction with cofactors and the basal transcription machinery. It is conceivable that interruption of any of these pathways could result in a defective HPA axis or glucocorticoid response leading either to a lack of glucocorticoid production or to glucocorticoid resistance and resultant enhanced autoimmune/inflammatory disease. Depending on the specific defect, patients might or might not respond to exogenous glucocorticoid therapy. For further reading on glucocorticoid resistance syndromes, see the recent review by Kino & Chrousos (161).

Disruptions in the HPA Axis or Glucocorticoid Response Leading to Glucocorticoid Sensitivity and Resistance In patient populations it is often difficult to dissect where a problem lies within the HPA axis, as many of its regulatory components cannot be directly

HPA AXIS

5 Feb 2002

7:56

Annu. Rev. Immunol. 2002.20:125-163. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

146

AR

AR152-06.tex

AR152-06.SGM

LaTeX2e(2001/05/10)

P1: GJC

WEBSTER ET AL.

measured in blood. Furthermore, inflammatory illness itself stimulates the HPA axis and alters its regulation (162, 163). Since CRH cannot be measured directly in peripheral venous blood, CRH responsiveness must be deduced by measuring levels of ACTH and cortisol, which can be directly measured in peripheral blood. A blunted HPA axis resulting in low levels of glucocorticoids or a low glucocorticoid response to HPA axis stimulation has been implicated in a number of inflammatory diseases, including rheumatoid arthritis (164–167), systemic lupus erythematosus (SLE) (168), Sjogren’s syndrome (169), allergic asthma and atopic skin disease (170), and chronic fatigue syndrome (171, 172). Patients with fibromyalgia have a low 24-h urinary free cortisol and a reduced response in the secretion of cortisol to HPA axis challenge (173). In patients with multiple sclerosis and a high plasma basal cortisol level, a normal HPA axis response to oCRH but a reduced HPA axis response to AVP were observed (150). Isolated glucocorticoid deficiency (IGD) is an autosomal recessive disorder, characterized by adrenocortical but not mineralocorticoid deficiency. Patients generally have low, undetectable cortisol levels that do not increase with treatment with exogenous ACTH. Also, the endogenous ACTH levels are high. Seventeen-point mutation and frameshifts in the receptor for ACTH have been identified that lead to this condition (174–176). Thus, this disease results because defective ACTH receptors render the adrenal gland incapable of sensing the high ACTH levels that continue to be secreted by the pituitary. Thus, despite high levels of ACTH, little or no cortisol is produced. Novel mutations in the ACTH receptor have also been described in a patient with ACTH hypersensitivity syndrome. In this case the patient had normal cortisol levels but low, undetectable ACTH levels (177). There have been some reports of patients with IGD developing autoimmune disease, such as organ-specific autoimmunity and autoimmune-mediated hypothroidism (178, 179).

ADRENAL GLANDS

CORTISOL BINDING GLOBULIN (CBG) The level of CBG in the blood limits the amount of free cortisol available. Changes in the expression of this protein or in its binding capacity could thus also affect the availability of cortisol. In New World primates, the lower levels of CBG and the lower affinity of cortisol for CBG have been associated with the higher circulating levels of glucocorticoids, which have been suggested to compensate for target organ resistance present in these mammals (180). In some patients with long-standing Crohn’s disease, a partial or complete resistance to steroids has been described; this resistance could be due to the increased expression of CBG in these patients, thereby limiting the bioavailability of glucocorticoids (181). 11β-HYDROXYSTEROID DEHYDROGENASE This enzyme catalyzes the conversion between an active glucocorticoid and inactive glucocorticoid (45) as described earlier. Thus, changes in this enzyme could result in differences in circulating or tissue glucocorticoid concentrations. In obese men, an impairment in the conversion of

5 Feb 2002

7:56

AR

AR152-06.tex

AR152-06.SGM

LaTeX2e(2001/05/10)

P1: GJC

NEUROENDOCRINE REGULATION OF IMMUNITY

147

Annu. Rev. Immunol. 2002.20:125-163. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

the inactive cortisone to the active cortisol was noted, indicating an impairment in the type I 11β-hydroxysteroid dehydrogenase and resulting in a drop in plasma cortisol levels (182). A decrease in 11β-hydroxysteroid dehydrogenase mRNA in ulcerative colitis has also been demonstrated (183). GLUCOCORTICOID RECEPTOR Glucocorticoid resistance has been associated with mutation of GR. This has been particularly studied in relation to familial glucocorticoid resistance, a hereditary condition caused by a mutation in the GR with associated decreased number of receptors, decreased affinity or stability of the receptor, and a decrease in translocation of the receptor to the nucleus. To date the molecular defects of four families and one sporadic case have been determined. These include three different point mutations in the ligand-binding domain, one in the hinge region, and a deletion in the ligand-binding domain. [For a review see Kino & Chrousos (161)]. A mutation in the GR may not be the only mechanism by which a change in this gene could affect the sensitivity to glucocorticoids. A polymorphism in codon 363 of GR has been described and associated with an increased sensitivity to glucocorticoids (184). On the other hand, five polymorphisms in GR (including the one in codon 363) have been described in a normal population, and they cannot be correlated with glucocorticoid resistance (185). Furthermore, there are patients with glucocorticoid resistance with no mutation detectable in GR, indicating that other defects in the steps in the pathway leading to gene activation by glucocorticoids could result in glucocorticoid resistance (186). The GR has several phosphorylation sites of which the function is unclear. Mutation of these phosphorylation sites results in reduced transactivation capability of GR of a minimal promoter and reduces the stability of the GR protein (187). Such changes as in the phosphorylation status of GR could have profound effects on GR function and be one method by which glucocorticoid resistance could occur. GRβ has been proposed to act as a negative regulator of GR function. Thus, variations in the level of GRα and GRβ could also be associated with apparent glucocorticoid resistance, both initial and acquired. Glucocorticoid-resistant asthma has been associated with a higher expression of GRβ on the peripheral blood lymphocytes (188, 189). GRβ expression is induced by cytokine treatment (190); thus as inflammatory disease progresses and cytokines are produced, induced GRβ expression could lead to exacerbation or development of glucocorticoid-resistant disease states (189). Increased expression of GRβ has been shown in mononuclear blood cells from patients with glucocorticoid-resistant ulcerative colitis compared to similar cells from patients with glucocorticoid-sensitive disease (191). A patient with generalized glucocorticoid resistance and chronic lymphoid leukemia has been described as having a decreased GR number and a reduced affinity for dexamethasone. This has been shown to be due not to a mutation in the GR but to a redistribution of GRα to GRβ with a reduced expression of GRα and an increased expression of GRβ (192). Finally, a polymorphism in exon 9β of the human GRβ gene has been found to be associated with rheumatoid arthritis. This polymorphism

5 Feb 2002

7:56

148

AR

AR152-06.tex

AR152-06.SGM

LaTeX2e(2001/05/10)

P1: GJC

WEBSTER ET AL.

Annu. Rev. Immunol. 2002.20:125-163. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

increases the stability of GRβ and thus alters glucocorticoid sensitivity (193). Taken together, these studies suggest that a variety of pre- and posttranslational changes in both forms of the GR are associated with autoimmune/inflammatory diseases. COFACTORS Although to date a mutation in a cofactor has not been found associated with glucocorticoid resistance, a recent study into the glucocorticoid hypersensitive state associated with HIV-1 infection is worthy of comment. This hypersensitive state is specific to the immune system, the brain, musculoskeletal system, and fat and liver with maintenance of an intact HPA system. A HIV-1 accessory protein, virion-associated protein (Vpr), is able to act as a transcriptional activator to aid the replication of the HIV-1 virus. This protein contains a nuclear receptor binding motif (LXXLL) that binds directly to GR and cooperatively enhances transcription via SRC1a and p300/CBP cofactors (161, 194). Recently, two sisters have been described with resistance to multiple steroids, and this may be due to a cofactor defect (195). TRANSPORT PROTEINS A group of ABC transmembrane transporters, MDR proteins, have been suggested to be involved in the active transport of glucocorticoids out of cells. Evidence exists that these proteins can transport glucocorticoids (196, 197; J. I. Webster, J. Carlstedt-Duke, unpublished data). These proteins are normally expressed in the blood-brain barrier, and knockout mice have been generated for the mouse mdr1a. In these knockout mice, an increase in the amount of dexamethasone present in the brain indicates that this protein is involved in the regulation of brain glucocorticoids (198, 199). These proteins have been identified through their involvement in the multidrug resistance phenotype (200). Recently, it was shown that patients with inflammatory bowel disease whom medical therapy had failed had an increase in the expression of the human MDR1 (ABCB1) (201). This raises the possibility that these proteins may be another component that could affect the effectiveness of glucocorticoid therapy by reducing the concentration of the hormone within the cells and thus reducing the ligand availability for the GR. The novel orphan receptor PXR is involved in the upregulation of MDR1 (202). Because PXR itself is activated by glucocorticoids among many other ligands (203), this could possibly represent another mechanism by which acquired glucocorticoid resistance may develop.

OTHER NEUROENDOCRINE FACTORS AFFECTING IMMUNE FUNCTION While this review has been focused on the HPA axis and glucocorticoid regulation of immunity and autoimmune/inflammatory disease, many other neural and neuroendocrine pathways connect with and regulate the immune system at multiple levels. These include the sympathetic nervous system and adrenergic factors and the peripheral nervous system and neuropeptides released by peripheral nerves.

5 Feb 2002

7:56

AR

AR152-06.tex

AR152-06.SGM

LaTeX2e(2001/05/10)

P1: GJC

NEUROENDOCRINE REGULATION OF IMMUNITY

149

Annu. Rev. Immunol. 2002.20:125-163. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

Adrenergic Factors The adrenergic sympathetic nervous system modulates local immune responses [for a comprehensive review of this topic, see Madden et al. (204)]. Noradrenergic sympathetic nerve fibers run from the CNS to primary and secondary lymphoid organs, such as the thymus, spleen, and lymph nodes. These nerve terminals make synaptic-like connections with neighboring immune cells releasing noradrenalin (205). The primary neurotransmitter released from sympathetic nerves, noradrenalin, exerts its effect at the target tissues through its receptor, the adrenergic receptor. Numerous cells of the immune system including lymphocytes and macrophages express adrenoreceptors. These are G-protein coupled receptors that can be divided into two subgroups—the α- and β-adrenergic receptors. These can be further subdivided into α 1-, α 2-, β 1-, β 2-, and β 3-adrenergic receptors. The most important receptor in terms of the immune system is the β 2-adrenergic receptor. Expression of α- and β-adrenergic receptors on T and B lymphocytes, neutrophils, mononuclear cells, and NK cells has been described. Activation of β 2-adrenergic receptors results in an increase in cyclic AMP concentrations, which can modulate cytokine expression, i.e., decreasing TNF-α and increasing IL-8 (94). It was originally thought that noradrenalin activated the β 2-adrenergic receptor resulting in suppressed lymphocyte function. However, more recent data suggest that it modulates B cell function to protect against or aid progression of immune disease. If α-adrenergic receptors are present on immune cells, then activation of these receptors will lead to the activation of a different signaling cascade and the activation of MAP kinases. Activation of this cascade has different effects on cellular activity of immune cells than does activation via the β2 adrenergic receptor (206–208). Furthermore, there appear to be regional difference in the effects of noradrenalin on immune function. Noradrenalin in the thymus is thought to play a role in modulation of thymocyte proliferation and differentiation, whereas in the spleen and lymph nodes it is thought to be involved in enhancement of the primary antibody response (205).

Sex Hormones The role of sex hormones, particularly estrogen, in the modulation of the immune system is an extensive and important area of research that is not reviewed here [for further reading, see Jansson & Holmdahl (209)]. It is noteworthy that there is a greatly increased risk—in the range of two- to tenfold higher female-tomale ratio—of developing many autoimmune/inflammatory diseases such as SLE, rheumatoid arthritis, and multiple sclerosis in women compared to men. Regulation of the immune system by estrogens is of particular importance during pregnancy. In this case a balance between glucocorticoid and estrogen regulation probably plays a role in suppression of the maternal immune system to prevent rejection of the fetus (209). Animal models have provided evidence for the importance of in vivo modulation of the immune system by the estrogen receptor (210, 211). There are two receptors for estrogens. Estrogen receptor α (ERα) was the first to be described (212). A second independent gene was identified in 1996 and termed

5 Feb 2002

7:56

Annu. Rev. Immunol. 2002.20:125-163. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

150

AR

AR152-06.tex

AR152-06.SGM

LaTeX2e(2001/05/10)

P1: GJC

WEBSTER ET AL.

ERβ (213). Knockout mice have shown that both estrogen receptors are important in a gender-specific manner for thymus development and atrophy (214). In addition, tamoxifen, an estrogen receptor antagonist, has some inhibitory effects on inflammation in LEW/N female rats (215). Sex hormone regulation of the immune system also varies with aging; for example, expression of IL-6 is under the control of sex hormones such as estrogen and testosterone. In postmenopausal women and postandropausal men, the loss of regulation of IL-6 results in increased concentrations of IL-6 that is associated with the increased occurrence of inflammatory diseases with old age such as rheumatoid arthritis, bowel disease, and osteoporosis (84).

CONCLUSION In summary there are multiple neuro-anatomical, hormonal, and molecular mechanisms by which the CNS regulates immune function and plays a role in susceptibility to and pathogenesis of autoimmune/inflammatory disease. We have focused here on the HPA axis and glucocorticoids, the final effector end point of the HPA axis in regulating immunity. It is clear that even in the case of a single hormone, there are many potential mechanisms at gene, protein, receptor, signaling, and cell function levels where dysregulation could lead to disease. Given the many different hormonal and nerve pathways that regulate immunity, each with its own specific molecular end points, the potential mechanisms for pathogenesis of autoimmune disease(s) resulting from disruptions in these interactions is large. Nonetheless, a thorough understanding at all levels of the means by which the CNS and immune systems communicate will provide many new insights into the bidirectional regulation of these systems and the disruptions in these communications that lead to disease. Ultimately this understanding will inform new avenues of therapy. ACKNOWLEDGMENTS The authors would like to thank Melanie Vacchio for critically reading the manuscript and for useful discussions. Visit the Annual Reviews home page at www.annualreviews.org

LITERATURE CITED 1. Hench PS, Kendall EC, Slocumb CH, Polley HF. 1950. Effects of cortisone acetate and primary ACTH on rheumatoid arthritis, rheumatic fever and certain other conditions. Arch. Intern. Med. 85:545– 666

2. Adcock IM. 2000. Molecular mechanisms of glucocorticosteroid actions. Pulmon. Pharmacol. Ther. 13:115–26 3. Barnes PJ, Adcock IM. 1998. Transcription factors and asthma. Eur. Resp. J. 12: 221–34

5 Feb 2002

7:56

AR

AR152-06.tex

AR152-06.SGM

LaTeX2e(2001/05/10)

P1: GJC

Annu. Rev. Immunol. 2002.20:125-163. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

NEUROENDOCRINE REGULATION OF IMMUNITY 4. Watkins LR, Maier SF, Goehler LE. 1995. Cytokine-to-brain communication: a review and analysis of alternative mechanisms. Life Sci. 57:1011–26 5. Dantzer R, Bluthe RM, Laye S, BretDibat JL, Parnet P, Kelley KW. 1998. Cytokines and sickness behavior. Ann. NY Acad. Sci. 840:586–90 6. Benveniste EN. 1998. Cytokine actions in the central nervous system. Cytokine Growth Factor Rev. 9:259–75 7. Hopkins SJ, Rothwell NJ. 1995. Cytokines and the nervous system. I. Expression and recognition. Trends Neurosci. 18:83–88 8. Brenneman DE, Schultzberg M, Bartfai T, Gozes I. 1992. Cytokine regulation of neuronal cell survival. J. Neurochem. 58: 454–60 9. Banks WA, Kastin AJ, Broadwell RD. 1995. Passage of cytokines across the blood-brain barrier. Neuroimmunomodulation 2:241–48 10. Dantzer R, Konsman JP, Bluthe RM, Kelley KW. 2000. Neural and humoral pathways of communication from the immune system to the brain: parallel or convergent? Autonomic Neurosci. 85:60– 65 11. Goehler LE, Gaykema RP, Hansen MK, Anderson K, Maier SF, Watkins LR. 2000. Vagal immune-to-brain communication: a visceral chemosensory pathway. Autonomic Neurosci. 85:49–59 12. Elmquist JK, Scammell TE, Saper CB. 1997. Mechanisms of CNS response to systemic immune challenge: the febrile response. Trends Neurosci. 20:565–70 13. Gaillard RC. 1998. Cytokines in the neuroendocrine system. Int. Rev. Immunol. 17:181–216 14. Mulla A, Buckingham JC. 1999. Regulation of the hypothalamo-pituitary-adrenal axis by cytokines. Baillieres Best Prac. Res. Clin. Endocrinol. Metab. 13:503– 21 15. Argo A, Stanisz AM. 1992. Are lymphocytes a target for substance P modula-

16.

17.

18.

19.

20.

21.

22.

23.

24.

151

tion in arthritis? Semin. Arthritis Rheum. 21:252–58 Crofford LJ, Sano H, Karalis K, Webster EL, Goldmuntz EA, Chrousos GP, Wilder RL. 1992. Local secretion of corticotrophin-releasing hormone in the joints of Lewis rats with inflammatory arthritis. J. Clin. Invest. 90:2555–64 Scott LV, Dinan TG. 1998. Vasopressin and the regulation of hypothalamicpituitary-adrenal axis function: implications for the pathophysiology of depression. Life Sci. 62:1985–98 Calogero AE, Gallucci WT, Gold PW, Chrousos GP. 1988. Multiple feedback regulatory loops upon rat hypothalamic corticotrophin-releasing hormone secretion. Potential clinical implications. J. Clin. Invest. 82:767–74 DeRijk RH, Petrides J, Deuster P, Gold PW, Sternberg EM. 1996. Changes in corticosteroid sensitivity of peripheral blood lymphocytes after strenuous exercise in humans. J. Clin. Endocrinol. Metab. 81:228–35 DeRijk R, Sternberg EM. 1997. Corticosterone resistance and disease. Ann. Med. 29:79–82 Singh A, Zelazowski EB, Petrides JS, Raybourne RB, Sternberg EM, Gold PW, Deuster PA. 1996. Lymphocyte subset responses to exercise and glucocorticoid suppression in healthy men. Med. Sci. Sports Exercise 28:822–28 Zelazowski EB, Singh A, Raybourne RB, Sternberg EM, Gold PW, Deuster PA. 1997. Lymphocyte subpopulation expression in women: effect of exercise and circadian rhythm. Med. Sci. Sports Exercise 29:467–73 Deuster PA, Zelazowski EB, Singh A, Sternberg EM. 1999. Expression of lymphocyte subsets after exercise and dexamethasone in high and low stress responders. Med. Sci. Sports Exerc. 31:1799– 806 Wilder RL, Calandra GB, Garvin AJ, Wright KD, Hansen CT. 1982. Strain and

5 Feb 2002

7:56

152

Annu. Rev. Immunol. 2002.20:125-163. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

25.

26.

27.

28.

29.

30.

31.

AR

AR152-06.tex

AR152-06.SGM

LaTeX2e(2001/05/10)

P1: GJC

WEBSTER ET AL. sex variation in the susceptibility to streptococcal cell wall-induced polyarthritis in the rat. Arthritis Rheum. 25:1064–72 Sternberg EM, Hill JM, Chrousos GP, Kamilaris T, Listwak SJ, Gold PW, Wilder RL. 1989. Inflammatory mediatorinduced hypothalamic-pituitary-adrenal axis activation is defective in streptococcal cell wall arthritis-susceptible Lewis rats. Proc. Natl. Acad. Sci. USA 86:2374– 78 Sternberg EM, Young WS III, Bernardini R, Calogero AE, Chrousos GP, Gold PW, Wilder RL. 1989. A central nervous system defect in biosynthesis of corticotrophin-releasing hormone is associated with susceptibility to streptococcal cell wall-induced arthritis in Lewis rats. Proc. Natl. Acad. Sci. USA 86:4771– 75 Dhabhar FS, McEwen BS, Spencer RL. 1993. Stress response, adrenal steroid receptor levels and corticosteroid-binding globulin levels—a comparison between Sprague-Dawley, Fischer 344 and Lewis rats. Brain Res. 616:89–98 Moncek F, Kvetnansky R, Jezova D. 2001. Differential responses to stress stimuli of Lewis and Fischer rats at the pituitary and adrenocortical level. Endocrinol. Regul. 35:35–41 Smith CC, Omelijaniuk RJ, Whitfield HJ Jr, Aksentijevich S, Fellows MQ, Zelazowski E, Gold PW, Sternberg EM. 1994. Differential mineralocorticoid (type 1) and glucocorticoid (type 2) receptor expression in Lewis and Fischer rats. Neuroimmunomodulation 1:66–73 Dhabhar FS, Miller AH, McEwen BS, Spencer RL. 1995. Differential activation of adrenal steroid receptors in neural and immune tissues of Sprague Dawley, Fischer 344, and Lewis rats. J. Neuroimmunol. 56:77–90 Edwards CK III, Yunger LM, Lorence RM, Dantzer R, Kelley KW. 1991. The pituitary gland is required for protection against lethal effects of Salmonella ty-

32.

33.

34.

35.

36.

37.

38.

39.

40.

phimurium. Proc. Natl. Acad. Sci. USA 88:2274–77 Ruzek MC, Pearce BD, Miller AH, Biron CA. 1999. Endogenous glucocorticoids protect against cytokine-mediated lethality during viral infection. J. Immunol. 162:3527–33 Misiewicz B, Poltorak M, Raybourne RB, Gomez M, Listwak S, Sternberg EM. 1997. Intracerebroventricular transplantation of embryonic neuronal tissue from inflammatory resistant into inflammatory susceptible rats suppresses specific components of inflammation. Exp. Neurol. 146:305–14 MacPhee IAM, Antoni FA, Mason DW. 1989. Spontaneous recovery of rats from experimental allergic encephalomyelitis is dependent on regulation of the immune system by endogenous adrenal corticosteroids. J. Exp. Med. 169:431–45 Rozlog LA, Kiecolt-Glaser JK, Marucha PT, Sheridan JF, Glaser R. 1999. Stress and immunity: implications for viral disease and wound healing. J. Periodontol. 70:786–92 Glaser R, Kiecolt-Glaser JK. 1998. Stress-associated immune modulation: relevance to viral infections and chronic fatigue syndrome. Am. J. Med. 105:35S– 42S Vedhara K, Cox NKM, Wilcock GK, Perks P, Hunt M, Anderson S, Lightman SL. 1999. Chronic stress in elderly carers of dementia patients and antibody response to influenza vaccination. Lancet 353:627–31 Friedl KE, Moore RJ, Hoyt RW, Marchitelli LJ, Matinez-Lopez LE, Askew EW. 2000. Endocrine markers of semistarvation in healthy lean men in a multistressor environment. J. Appl. Physiol. 88:1820–30 de Kloet ER, Vreugdenhil E, Oitzl MS, Jo¨els M. 1998. Brain corticosteroid receptor balance in health and disease. Endocrine Rev. 19:269–301 Deleted in proof

5 Feb 2002

7:56

AR

AR152-06.tex

AR152-06.SGM

LaTeX2e(2001/05/10)

P1: GJC

Annu. Rev. Immunol. 2002.20:125-163. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

NEUROENDOCRINE REGULATION OF IMMUNITY 41. Funder JW. 1997. Glucocorticoid and mineralocorticoid receptors: biology and clinical relevance. Annu. Rev. Med. 48: 231–40 42. Liu W, Wang J, Sauter NK, Pearce D. 1995. Steroid receptor heterodimerization demonstrated in vitro and in vivo. Proc. Natl. Acad. Sci. USA 92:12480–84 43. Trapp T, Holsboer F. 1996. Heterodimerization between mineralocorticoid and glucocorticoid receptors increases the functional diversity of corticosteroid action. Trends Pharmacol. Sci. 17:145–49 44. de Kloet ER. 2000. Stress in the brain. Eur. J. Pharmacol. 405:187–98 45. Seckl JR, Walker BR. 2001. Minireview: 11β-hydroxysteroid dehydrogenase type 1—a tissue-specific amplifier of glucocorticoid action. Endocrinology 142:1371– 76 46. Nuclear Receptors Nomenclature Committee. 1999. A unified nomenclature system for the nuclear receptor superfamily. Cell 97:161–63 47. Evans RM. 1988. The steroid and thyroid hormone receptor superfamily. Science 240:889–95 48. McKay LI, Cidlowski JA. 1999. Molecular control of immune/inflammatory responses: interactions between nuclear factor-κB and steroid receptor-signaling pathways. Endocrine Rev. 20:435–59 49. Enc´ıo IJ, Detera-Wadleigh SD. 1991. The genomic structure of the human glucocorticoid receptor. J. Biol. Chem. 266:7182– 88 50. Oakley RH, Jewell CM, Yudt MR, Bofetiado DM, Cidlowski JA. 1999. The dominant negative activity of the human glucocorticoid receptor β isoform. Specificity and mechanisms of action. J. Biol. Chem. 274:27857–66 51. Oakley RH, Sar M, Cidlowski JA. 1996. The human glucocorticoid receptor β isoform. Expression, biochemical properties and putative function. J. Biol. Chem. 271:9550–59 52. Vottero A, Chrousos GP. 1999. Gluco-

53.

54.

55.

56.

57.

58.

59.

60.

61.

153

corticoid receptor β: View I. Trends Endocrinol. Metab. 10:333–38 Hecht K, Carlstedt-Duke J, Stierna P, ˚ Br¨onneg˚ard M, WikGustafsson J-A, str¨om A-C. 1997. Evidence that the βisoform of the human glucocorticoid receptor does not act as a physiologically significant repressor. J. Biol. Chem. 272:26659–64 Brogan IJ, Murray IA, Cerillo G, Needham M, White A, Davis JRE. 1999. Interaction of glucocorticoid receptor isoforms with transcription factors AP-1 and NF-κB: lack of effect of glucocorticoid receptor β. Mol. Cell. Endocrinol. 157:95– 104 Carlstedt-Duke J. 1999. Glucocorticoid receptor β: View II. Trends Endocrinol. Metab. 10:339–42 Sch¨ule R, Rangarajan P, Kliewer S, Ransone LJ, Bolado J, Yang N, Verma IM, Evans RM. 1990. Functional antagonism between oncoprotein c-Jun and the glucocorticoid receptor. Cell 62:1217–26 Adcock IM, Brown CR, Shirasaki H, Barnes PJ. 1994. Effects of dexamethasone on cytokine and phorbol ester stimulated c-Fos and c-Jun DNA binding and gene expression in human lung. Eur. Resp. J. 7:2117–23 Pearce D, Matsui W, Miner JN, Yamamoto KR. 1998. Glucocorticoid receptor transcriptional activity determined by spacing of receptor and nonreceptor DNA sites. J. Biol. Chem. 273:30081– 85 Pearce D, Yamamoto KR. 1993. Mineralocorticoid and glucocorticoid receptor activities distinguished by nonreceptor factors at a composite response element. Science 259:1161–65 Malkoski SP, Dorin RI. 1999. Composite glucocorticoid regulation at a functionally defined negative glucocorticoid response element of the human corticotrophinreleasing hormone gene. Mol. Endocrinol. 13:1629–44 Baldwin AS Jr. 1996. The NF-κB and IκB

5 Feb 2002

7:56

154

62.

Annu. Rev. Immunol. 2002.20:125-163. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

63.

64.

65.

66.

67.

68.

69.

AR

AR152-06.tex

AR152-06.SGM

LaTeX2e(2001/05/10)

P1: GJC

WEBSTER ET AL. proteins: new discoveries and insights. Annu. Rev. Immunol. 14:649–81 Ghosh S, May MJ, Kopp EB. 1998. NFκB and Rel proteins: Evolutionarily conserved mediators of immune responses. Annu. Rev. Immunol. 16:225–60 van der Burg B, Liden J, Okret S, Delaunay F, Wissink S, van der Saag PT, ˚ 1997. Nuclear factor-κ Gustafsson J-A. B repression in antiinflammation and immunosuppression by glucocorticoids. Trends Endocinol. Metab. 8:152–57 Ramdas J, Harmon JM. 1998. Glucocorticoid-induced apoptosis and regulation of NF-κB activity in human leukemic T cells. Endocrinology 139:3813–21 Ray A, Prefontaine KE. 1994. Physical association and functional antagonism between the p65 subunit of transcription factor NF-κB and the glucocorticoid receptor. Proc. Natl. Acad. Sci. USA 91: 752–56 Caldenhoven E, Liden J, Wissink S, Van de Stolpe A, Raaijmakers J, Koender˚ Van man L, Okret S, Gustafsson J-A, der Saag PT. 1995. Negative cross-talk between RelA and the glucocorticoid receptor: a possible mechanism for the antiinflammatory action of glucocorticoids. Mol. Endocrinol. 9:401–12 Scheinman RI, Gualberto A, Jewell CM, Cidlowski JA, Baldwin AS Jr. 1995. Characterization of mechanisms involved in transrepression of NF-κB by activated glucocorticoid receptors. Mol. Cell. Biol. 15:943–53 Kleinert H, Euchenhofer C, Ihrig-Biedert I, F¨orstermann U. 1996. Glucocorticoids inhibit the induction of nitric oxide synthase II by down-regulating cytokineinduced activity of transcription factor nuclear factor-κB. Mol. Pharmacol. 49:15– 21 De Bosscher K, Schmitz ML, Vanden Berghe W, Plaisance S, Fiers W, Haegeman G. 1997. Glucocorticoid-mediated repression of nuclear factor-κB-dependent transcription involved direct inter-

70.

71.

72.

73.

74.

75.

76.

77.

ference with transactivation. Proc. Natl. Acad. Sci. USA 94:13504–9 Ray KP, Farrow S, Daly M, Talabot F, Searle N. 1997. Induction of the Eselectin promoter by interleukin-1 and tumour necrosis factor α, and inhibition by glucocorticoids. Biochem. J. 328:707–15 Liden J, Rafter I, Truss M, Gustafsson J˚ Okret S. 2000. Glucocorticoid effects A, on NF-κB binding in the transcription of the ICAM-1 gene. Biochem. Biophys. Res. Commun. 273:1008–14 Wissink S, van Heerde EC, van der Burg B, van der Saag PT. 1998. A dual mechanism mediates repression of NF-κB activity by glucocorticoids. Mol. Endocrinol. 12:355–63 Liden J, Delaunay F, Rafter I, Gustafsson ˚ Okret S. 1997. A new function for J-A, the C-terminal zinc finger of the glucocorticoid receptor. Repression of RelA transactivation. J. Biol. Chem. 272:21467– 72 Wissink S, van de Stolpe A, Caldenhoven E, Koenderman L, van der Saag PT. 1997. NF-kB/Rel family members regulating the ICAM-1 promoter in monocytic THP-1 cells. Immunobiology 198:50– 64 Sheppard K-A, Phelps KM, Williams AJ, Thanos D, Glass K, Rosenfeld MG, Gerritsen ME, Collins T. 1998. Nuclear integration of glucocorticoid receptor and nuclear factor-κB signaling by CREB-binding protein and steroid receptor coactivator-1. J. Biol. Chem. 273: 29291–94 McKay LI, Cidlowski JA. 2000. CBP (CREB binding protein) integrates NF-κB (nuclear factor-κB) and glucocorticoid receptor physical interaction and antagonism. Mol. Endocrinol. 14:1222–34 Doucas V, Shi Y, Miyamoto S, West A, Verma I, Evans RM. 2000. Cytoplasmic catalytic subunit of protein kinase A mediates cross-repression by NF-κB and the glucocorticoid receptor. Proc. Natl. Acad. Sci. USA 97:11893–98

5 Feb 2002

7:56

AR

AR152-06.tex

AR152-06.SGM

LaTeX2e(2001/05/10)

P1: GJC

Annu. Rev. Immunol. 2002.20:125-163. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

NEUROENDOCRINE REGULATION OF IMMUNITY 78. Barnes PJ. 1998. Anti-inflammatory actions of glucocorticoids: molecular mechanisms. Clin. Sci. 94:557–72 79. Lee SW, Tsou A-P, Chan H, Thomas J, Petrie K, Eugui EM, Allison AC. 1988. Glucocorticoids selectively inhibit the transcription of the interleukin 1β gene and decrease the stability of interleukin 1β mRNA. Proc. Natl. Acad. Sci. USA 85: 1204–8 80. Groujon E, Lay´e S, Parnet P, Dantzer R. 1997. Regulation of cytokine gene expression in the central nervous system by glucocorticoids: mechanisms and functional consequences. Psychoneuroendocrinology 22:S75–S80 81. Schmidt M, Pauels H-G, L¨ugering N, L¨ugering A, Domschke W, Kucharzik T. 1999. Glucocorticoids induce apoptosis in human monocytes: Potential role of IL-1β. J. Immunol. 163:3484–90 82. Kunicka JE, Talle MA, Denhardt GH, Brown M, Prince LA, Goldstein G. 1993. Immunosuppression by glucocorticoids: Inhibition of production of multiple lymphokines by in vivo administration of dexamethasone. Cell. Immunol. 149:39– 49 83. Colotta F, Re F, Muzio M, Bertini R, Polentarutti N, Sironi M, Giri JG, Dower SK, Sims JE, Mantovani A. 1993. Interleukin-1 type II receptor: A decoy target for IL-1 that is regulated by IL-4. Science 261:472–75 84. Vanden Berghe W, Vermeulen L, De Wilde G, De Bosscher K, Boone E, Haegeman G. 2000. Signal transduction by tumor necrosis factor and gene regulation of the inflammatory cytokine interleukin-6. Biochem. Pharmacol. 60: 1185–95 85. Wang J, Zhu Z, Nolfo R, Elias JA. 1999. Dexamethasone regulation of lung epithelial cell and fibroblast interleukin-11 production. Am. J. Physiol. Lung Cell. Mol. Physiol. 276:L175–L85 86. Steer JH, Ma DTS, Dusci L, Garas G, Pederson KE, Joyce DA. 1998. Al-

87.

88.

89.

90.

91.

92.

93.

155

tered leucocyte trafficking and suppressed tumour necrosis factor -α release from peripheral blood monocytes after intraarticular glucocorticoid treatment. Ann. Rheu. Dis. 57:732–37 Steer JH, Kroeger KM, Abraham LJ, Joyce DA. 2000. Glucocorticoids suppress tumor necrosis factor-α expression by human monocytic THP-1 cells by suppressing transactivation through adjacent NF-κB and c-Jun-activating transcription factor-2 binding sites in the promoter. J. Biol. Chem. 275:18432–40 Joyce DA, Steer JH, Abraham LJ. 1997. Glucocorticoid modulation of human monocyte/macrophage function: Control of TNF-α secretion. Inflamm. Res. 46: 447–51 Moynihan JA, Callahan TA, Kelley SP, Campbell LM. 1998. Adrenal hormone modulation of type 1 and type 2 cytokine production by spleen cells: Dexamethasone and dehydroepiandrosterone suppress interleukin-2, interleukin-4, and interferon-γ production in vitro. Cell. Immunol. 184:58–64 Verhoef CM, van Roon JAG, Vianen ME, Lafeber FPJG, Bijlsma JWJ. 1999. The immune suppressive effect of dexamethasone in rheumatoid arthritis is accompanied by upregulation of interleukin 10 and by differential changes in interferon γ and interleukin 4 production. Ann. Rheu. Dis. 58:49–54 Chang MM-J, Juarez M, Hyde DM, Wu R. 2001. Mechanism of dexamethasonemediated interleukin-8 gene suppression in cultured airway epithelial cells. Am. J. Physiol. Lung Cell. Mol. Physiol. 280:L107–L15 Wu C-Y, Wang K, McDyer JF, Seder RA. 1998. Prostaglandin E2 and dexamethasone inhibit IL-12 receptor expression and IL-12 responsiveness. J. Immunol. 161:2723–30 Franchimont D, Galon J, Gadina M, Visconti R, Zhou Y-J, Aringer M, Frucht DM, Chrousos GP, O’Shea JJ. 2000.

5 Feb 2002

7:56

156

Annu. Rev. Immunol. 2002.20:125-163. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

94.

95.

96.

97.

97a.

98.

99.

100.

101.

AR

AR152-06.tex

AR152-06.SGM

LaTeX2e(2001/05/10)

P1: GJC

WEBSTER ET AL. Inhibition of Th1 immune response by glucocorticoids: dexamethasone selectively inhibits IL-12-induced Stat4 phosphorylation in T lymphocytes. J. Immunol. 164:1768–74 Elenkov IJ, Chrousos GP. 1999. Stress hormones, Th1/Th2 patterns, pro/antiinflammatory cytokines and susceptibility to disease. Trends Endocrinol. Metab. 10:359–68 Bianchi M, Meng C, Ivashkiv LB. 2000. Inhibition of IL-2-induced Jak-STAT signaling by glucocorticoids. Proc. Natl. Acad. Sci. USA 97:9573–78 Adkins KK, Levan TD, Miesfeld RL, Bloom JW. 1998. Glucocorticoid regulation of GM-CSF: evidence for transcriptional mechanisms in airway epithelial cells. Am. J. Physiol. Lung Cell. Mol. Physiol. 275:L372–L78 Hodge S, Hodge G, Flower R, Han P. 1999. Methyl-prednisolone up-regulates monocyte interleukin-10 production in stimulated whole blood. Scand. J. Immunol. 49:548–53 Visser J, van Boxel-Dezaire A, Methorst D, Brunt T, de Kloet ER, Nagelkerken L. 1998. Differential regulation of interleukin-10 (IL-10) and IL-12 by glucocorticoids in vitro. Blood 91:4255–64 Yoshikawa H, Nakajima Y, Tasaka K. 1999. Glucocorticoid suppresses autocrine survival of mast cells by inhibiting IL-4 production and ICAM-1 expression. J. Immunol. 162:6162–70 Sewell WA, Scurr LL, Orphanides H, Kinder S, Ludowyke RI. 1998. Induction of interleukin-4 and interleukin-5 expression in mast cells is inhibited by glucocorticoids. Clin. Diagn. Lab. Immunol. 5:18– 23 Daynes RA, Araneo BA. 1989. Contrasting effects of glucocorticoids on the capacity of T cells to produce the growth factors interleukin 2 and interleukin 4. Eur. J. Immunol. 19:2319–25 Daynes RA, Araneo BA, Dowell TA, Huang K, Dudley D. 1990. Regula-

102.

103.

104.

105.

106.

107.

108.

tion of murine lymphokine production in vivo. III. The lymphoid tissue microenvironment exerts regulatory influences over T helper cell function. J. Exp. Med. 171:979–96 Cronstein BN, Kimmel SC, Levin RI, Martiniuk F, Weissmann G. 1992. A mechanism for the antiinflammatory effects of corticosteroids: the glucocorticoid receptor regulates leukocyte adhesion to endothelial cells and expression of endothelial-leukocyte adhesion molecule 1 and intercellular adhesion molecule 1. Proc. Natl. Acad. Sci. USA 89:9991–95 Atsuta J, Plitt J, Bochner BS, Schleimer RP. 1999. Inhibition of VCAM-1 expression in human bronchial epithelial cells by glucocorticoids. Am. J. Resp. Cell Mol. Biol. 20:643–50 Nakagawa M, Bondy GP, Waisman D, Minshall D, Hogg JC, van Eeden SF. 1999. The effect of glucocorticoids on the expression of L-selectin on polymorphonuclear leukocyte. Blood 93:2730–37 Ohtsuka T, Kubota A, Hirano T, Watanabe K, Yoshida H, Tsurufuji M, Iizuka Y, Konishi K, Tsurufuji S. 1996. Glucocorticoid-mediated gene suppression of rat cytokine-induced neutrophil chemoattractant CINC/gro, a member of the interleukin-8 family, through impairment of NF-κB activation. J. Biol. Chem. 271:1651–59 Richards DF, Fernandez M, Caulfield J, Hawrylowicz CM. 2000. Glucocorticoids drive human CD8+ T cell differentiation towards a phenotype with high IL-10 and reduced IL-4, IL-5 and IL-13 production. Eur. J. Immunol. 30:2344–54 Mori A, Kaminuma O, Suko M, Inoue S, Ohmura T, Hoshino A, Asakura Y, Miyazawa K, Yokota T, Okumura Y, Ito K, Okudaira H. 1997. Two distinct pathways of interleukin-5 synthesis in allergen-specific human T-cell clones are suppressed by glucocorticoids. Blood 89:2891–2900 Pype JL, Dupont LJ, Menten P, Van

5 Feb 2002

7:56

AR

AR152-06.tex

AR152-06.SGM

LaTeX2e(2001/05/10)

P1: GJC

NEUROENDOCRINE REGULATION OF IMMUNITY

Annu. Rev. Immunol. 2002.20:125-163. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

109.

110.

111.

112.

113.

114.

115.

116.

Coillie E, Opdenakker G, Van Damme J, Chung KF, Demedts MG, Verleden GM. 1999. Expression of monocyte chemotactic protein (MCP)-1, MCP-2, and MCP3 by human airway smooth-muscle cells. Am. J. Resp. Cell Mol. Biol. 21:528–36 Wingett D, Forcier K, Nielson CP. 1996. Glucocorticoid-mediated inhibition of RANTES expression in human T lymphocytes. FEBS Lett. 398:308–11 Poon M, Liu B, Taubman MB. 1999. Identification of a novel dexamethasonesensitive RNA-destabilizing region on rat monocyte chemoattractant protein 1 mRNA. Mol. Cell. Biol. 19:6471–78 Lilly CM, Nakamura H, Kesselman H, Nagler-Anderson C, Asano K, GarciaZepeda EA, Rothenberg ME, Drazen JM, Luster AD. 1997. Expression of eotaxin by human lung epithelial cells. Induction by cytokines and inhibition by glucocorticoids. J. Clin. Invest. 99:1767–73 Yamamoto K, Arakawa T, Ueda N, Yamamoto S. 1995. Transcriptional roles of nuclear factor κB and nuclear factorinterleukin-6 in the tumor necrosis factor α-dependent induction of cylooxygenase2 in MC3T3-E1 cells. J. Biol. Chem. 270:31315–20 Perkins DJ, Kniss DA. 1997. Tumor necrosis factor-α promotes sustained cyclooxygenase-2 expression: attenuation by dexamethasone and NSAIDs. Prostaglandins 54:727–43 Newton R, Kuitert LM, Slater DM, Adcock IM, Barnes PJ. 1997. Cytokine induction of cytosolic phospholipase A2 and cyclooxygenase-2 mRNA is suppressed by glucocorticoids in human epithelial cells. Life Sci. 60:67–78 Dolan-O’Keefe M, Nick HS. 1999. Inhibition of cytoplasmic phospholipase A2 expression by glucocorticoid in rat intestinal epithelial cells. Gastroenterology 116:855–64 de Vera ME, Taylor BS, Wang Q, Shapiro RA, Billiar TR, Geller DA. 1997. Dexamethasone suppresses iNOS gene expres-

117.

118.

119.

120.

121.

122.

123.

124.

125.

126.

157

sion by upregulating I-κBα and inhibiting NF-κB. Am. J. Physiol. 273:G1290– G96 Buckingham JC, Flower RJ. 1997. Lipocortin 1: a second messenger of glucocorticoid action in the hypothalamopituitary-adrenocortical axis. Mol. Med. Today 3:296–302 Collins S, Caron MG, Lefkowitz RJ. 1988. Beta-adrenergic receptors in hamster smooth muscle cells are transcriptionally regulated by glucocorticoids. J. Biol. Chem. 263:9067–70 Mak JC, Nishikawa M, Barnes PJ. 1995. Glucocorticoids increase beta 2adrenergic receptor transcription in human lung. Am. J. Physiol. 268:L41–L46 Ihara H, Nakanishi S. 1990. Selective inhibition of expression of the substance P receptor mRNA in pancreatic acinar AR42J cells by glucocorticoids. J. Biol. Chem. 265:22441–45 Katsunuma T, Mak JC, Barnes PJ. 1998. Glucocorticoids reduce tachykinin NK2 receptor expression in bovine tracheal smooth muscle. Eur. J. Pharmacol. 344:99–106 Russo-Marie F. 1992. Macrophages and the glucocorticoids. J. Neuroimmunol. 40:281–86 O’Connor TM, O’Halloran DJ, Shanahan F. 2000. The stress response and the hypothalamic-pituitary-adrenal axis: from molecule to melancholia. Q. J. Med. 93:323–33 Goulding NJ, Euzger HS, Butt SK, Perretti M. 1998. Novel pathways for glucocorticoid effects on neutrophils in chronic inflammation. Inflamm. Res. 47:S158– S65 Nelms K, Keegan AD, Zamorano J, Ryan JJ, Paul WE. 1999. The IL-4 receptor: signaling mechanisms and biologic functions. Annu. Rev. Immunol. 17:701–38 Franchimont D, Louis E, Dewe W, Martens H, Vrindts-Gevaert Y, de Groote D, Belaiche J, Greenen V. 1998. Effects of dexamethasone on the profile of

5 Feb 2002

7:56

158

127.

Annu. Rev. Immunol. 2002.20:125-163. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

128.

129.

130.

131.

132.

133.

134.

135.

AR

AR152-06.tex

AR152-06.SGM

LaTeX2e(2001/05/10)

P1: GJC

WEBSTER ET AL. cytokine secretion in human whole blood cell cultures. Regl. Pept. 73:59–65 Agarwal SK, Marshall GD Jr. 2001. Dexamethasone promotes type 2 cytokine production primarily through inhibition of type 1 cytokines. J. Interferon Cytokine Res. 21:147–55 Tamada K, Harada M, Abe K, Li T, Nomoto K. 1998. IL-4-producing NK1.1+ T cells are resistant to glucocorticoidinduced apoptosis: Implications for the Th1/Th2 balance. J. Immunol. 161:1239– 47 Elenkov IJ, Hoffman J, Wilder RL. 1997. Does differential neuroendocrine control of cytokine production govern the expression of autoimmune diseases in pregnancy and the postpartum period? Mol. Med. Today 3:379–83 Schleimer RP, Bochner BS. 1994. The effects of glucocorticoids on human eosinophils. J. Allergy Clin. Immunol. 94:1202–13 Hirai K, Miyamasu M, Takaishi T, Morita Y. 1997. Regulation of the function of eosinophils and basophils. Crit. Rev. Immunol. 17:325–52 Cohan VL, Undem BJ, Fox CC, Adkinson NFJ, Lichtenstein LM, Schleimer RP. 1989. Dexamethasone does not inhibit the release of mediators from human mast cells residing in airway, intestine, or skin. Am. Rev. Respir. Dis. 140:951– 54 Laitinen LA, Laitinen ATH. 1992. A comparative study of the effects of an inhaled corticosteroid, budesonide, and a beta 2-agonist, terbutaline, on airway inflammation in newly diagnosed asthma: a randomized, double-blind, parallel-group controlled trial. J. Allergy Clin. Immunol. 90:32–42 Nelson DJ, McWilliam AS, Haining S, Holt PG. 1995. Modulation of airway intraepithelial dendritic cells following exposure to steroids. Am. J. Resp. Crit. Care Med. 151:475–81 Holm AF, Fokkens WJ, Godthelp T, Mul-

136.

137.

138.

139.

140.

141.

142.

143.

144.

der PG, Vroom TM, Rijntes E. 1995. Effect of 3 months’ nasal steroid therapy on nasal T cells and Langerhans cells in patients suffering from allergic rhinitis. Allergy 50:402–9 Boschetto P, Rogers DF, Fabbri LM, Barnes PJ. 1991. Corticosteroid inhibition of airway microvascular leakage. Am. Rev. Respir. Dis. 143:605–9 Adcock IM, Gilbey T, Gelder CM, Chung KF, Barnes PJ. 1996. Glucocorticoid receptor localization in normal and asthmatic lung. Am. J. Resp. Crit. Care Med. 154:771–82 Kai H, Yoshitake K, Hisatsune A, Kido T, Isohama Y, Takahama K, Miyata T. 1996. Dexamethasone suppresses mucus production and MUC-2 and MUC-5AC gene expression by NCI-H292 cells. Am. J. Physiol. 271: L484–L488 Shimura S, Sasaki T, Ikeda K, Yamaguchi K, Sasaki H, Takishima T. 1990. Direct inhibitory action of glucocorticoid on glycoconjugates secretion from airway submucosal glands. Am. Rev. Respir. Dis. 141:1044–49 King KL, Cidlowski JA. 1998. Cell cycle regulation and apoptosis. Annu. Rev. Physiol. 60:601–17 Planey SL, Litwack G. 2000. Glucocorticoid-induced apoptosis in lymphocytes. Biochem. Biophys. Res. Commun. 279:307–12 Samuelsson MKR, Pazirandeh A, Davani B, Okret S. 1999. P57kip2, a glucocorticoid-induced inhibitor of cell cycle progression in HeLa cells. Mol. Endocrinol. 13:1811–22 Schmidt M, L¨ugering N, L¨ugering A, Pauels H-G, Schulze-Osthoff K, Domschke W, Kucharzik T. 2001. Role of the CD95/CD95 ligand system in glucocorticoid-induced monocyte apoptosis. J. Immunol. 166:1344–51 Ashwell JD, Lu FWM, Vacchio MS. 2000. Glucocorticoids in T cell development and function. Annu. Rev. Immunol. 18:309–45

5 Feb 2002

7:56

AR

AR152-06.tex

AR152-06.SGM

LaTeX2e(2001/05/10)

P1: GJC

Annu. Rev. Immunol. 2002.20:125-163. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

NEUROENDOCRINE REGULATION OF IMMUNITY 145. Pazirandeh A, Xue Y, Rafter I, Sj¨ovall J, Jondal M, Okret S. 1999. Paracrine glucocorticoid activity produced by mouse thymic epithelial cells. FASEB J. 13:893–901 146. Vacchio MS, Ashwell JD, King LB. 1998. A positive role for thymus-derived steroids in formation of the T-cell repertoire. Ann. NY Acad. Sci. 840:317–27 147. Lechner O, Wiegers GJ, Oliveira-dosSantos AJ, Dietrich H, Recheis H, Waterman M, Boyd R, Wick G. 2000. Glucocorticoid production in the murine thymus. Eur. J. Immunol. 30:337–46 148. Vacchio MS, Ashwell JD. 1997. Thymus-derived glucocorticoids regulate antigen-specific positive selection. J. Exp. Med. 185:2033–38 149. Vacchio MS, Lee JJY, Ashwell JD. 1999. Thymus-derived glucocorticoids set the thresholds for thymocyte selection by inhibiting TCR-mediated thymocyte activation. J. Immunol. 163:1327– 33 150. Michelson D, Stone L, Galliven E, Magiakou MA, Chrousos GP, Sternberg EM, Gold PW. 1994. Multiple sclerosis is associated with alterations in hypothalamic-pituitary-adrenal axis function. J. Clin. Endocrinol. Metab. 79:848–53 151. Morale MC, Batticane N, Gallo F, Barden N, Marchetti B. 1995. Disruption of hypothalamic-pituitary-adrenocortical system in transgenic mice expressing type II glucocorticoid receptor antisense ribonucleic acid permanently impairs T cell function: Effects on T cell trafficking and T cell responsiveness during postnatal development. Endocrinology 136:3949–60 152. Reichardt HM, Kaestner KH, Tuckermann J, Kretz O, Wessely O, Bock R, Gass P, Schmid W, Herrlich P, Angel P, Sch¨utz G. 1998. DNA binding of the glucocorticoid receptor is not essential for survival. Cell 93:531–41 153. Godfrey DI, Purton JF, Boyd RL,

154.

155.

156.

157.

158.

158a.

158b.

158c.

159.

159

Cole TL. 2000. Stress-free T-cell development: Glucocorticoids are not obligatory. Immunol. Today 21:606–11 Purton JF, Boyd RL, Cole TJ, Godfrey DI. 2000. Intrathymic T cell development and selection proceeds normally in the absence of glucocorticoid receptor signaling. Immunity 13:179–86 Wick G, Sgonc R, Lechner O. 1998. Neuroendocrine-immune disturbances in animal models with spontaneous autoimmune diseases. Ann. NY Acad. Sci. 840:591–98 Hu Y, Dietrich H, Herold M, Heinrich PC, Wick G. 1993. Disturbed immunoendocrine communication via the hypothalamo-pituitary-adrenal axis in autoimmune disease. Int. Arch. Allergy Immunol. 102:232–41 Lechner O, Hu Y, Jafarian-Tehrani M, Dietrich H, Schwartz S, Herold M, Haour F, Wick G. 1996. Disturbed immunoendocrine communication via the hypothalamo-pituitary-adrenal axis in murine lupus. Brain Behav. Immun. 10:337–50 Jafarian-Tehrani M, Sternberg EM. 1999. Animal models of neuroimmune interaction in inflammatory diseases. J. Neuroimmunol. 100:13–20 Wakeland EK, Wandstrat AE, Liu K, Morel L. 1999. Genetic dissection of systemic lupus erythematosus. Curr. Opin. Immunol. 11:701–7 Remmers EF, Longman RE, Du Y, O’Hare A, Cannon GW, Griffiths MM, Wilder RL. 1996. A genome scan localizes five non-MHC loci controlling collagen-induced arthritis in rates. Nat. Genet. 14:82–85 Listwak S, Barrientos RM, Koike G, Ghosh S, Gomez M, Misiewica B, Sternberg EM. 1999. Identification of a novel inflammation-protective locus in the Fischer rat. Mamm. Genome 10:362– 65 Petrides JS, Gold PW, Mueller GP, Singh A, Stratakis C, Chrousos GP,

5 Feb 2002

7:56

160

Annu. Rev. Immunol. 2002.20:125-163. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

160.

161.

162.

163.

164.

165.

166.

AR

AR152-06.tex

AR152-06.SGM

LaTeX2e(2001/05/10)

P1: GJC

WEBSTER ET AL. Deuster PA. 1997. Marked differences in functioning of the hypothalamicpituitary-adrenal axis between groups of men. J. Appl. Physiol. 82:1979–88 Huizenga NATM, Koper JW, de Lange P, Pols HAP, Stolk RP, Grobbee DE, de Jong FH, Lamberts SWJ. 1998. Interperson variability but intraperson stability of baseline plasma cortisol concentrations, and its relation to feedback sensitivity of the hypothalamo-pituitaryadrenal axis to a low dose of dexamethasone in elderly individuals. J. Clin. Endocrinol. Metab. 83:47–54 Kino T, Chrousos GP. 2001. Glucocorticoid and mineralocorticoid resistance/hypersensitivity syndromes. J. Endocrinol. 169:437–45 Schmidt ED, Janszen AW, Wouterlood FG, Tilders FJ. 1995. Interleukin-1-induced long-lasting changes in hypothalamic corticotrophin-releasing hormone (CRH)–neurons and hyperresponsiveness of the hypothalamus-pituitaryadrenal axis. J. Neruosci. 15:7417–26 Chowdry HS, Larsen PJ, Harbuz MS, Jessop DS, Aguilera G, Eckland DJ, Lightman SL. 1995. Evidence for arginine vasopressin as the primary activator of the HPA axis during adjuvant-induced arthritis. Br. J. Pharmacol. 116:2417–24 Cash JM, Crofford LJ, Gallucci WT, Sternberg EM, Gold PW, Chrousos GP, Wilder RL. 1992. Pituitary-adrenal axis responsiveness to ovine corticotrophin releasing hormone in patients with rheumatoid arthritis treated with low dose prednisolone. J. Rheumatol. 19:1692–96 Chikanza IC, Petrou P, Kingsley G, Chrousos GP, Panayi GS. 1992. Defective hypothalamic response to immune and inflammatory stimuli in patients with rheumatoid arthritis. Arthritis Rheum. 35:1281–88 Cutolo M, Foppiani L, Perete C, Ballarino P, Sulli A, Villaggio B, Seriolo B, Giusti M, Accardo S. 1999.

167.

168.

169.

170.

171.

172.

173.

174.

Hypothalamic-pituitary-adrenocortical axis function in premenopausal women with rheumatoid arthritis not treated with glucocorticoids. J. Rheumatol. 26: 282–88 Gutierrez MA, Garcia ME, Rodriguez JA, Mardonez G, Jacobelli S, Rivero S. 1999. Hypothalamic-pituitary-adrenal axis in patients with active rheumatoid arthritis: a controlled study using insulin hypoglycemia stress test and prolactin stimulation. J. Rheumatol. 26:277–81 Gutierrez MA, Garcia ME, Rodriguez JA, Rivero S, Jacobelli S. 1998. Hypothalamic-pituitary-adrenal axis function and prolactin secretion in systemic lupus erythematosus. Lupus 7:404–8 Johnson EO, Vlachoyiannopoulos PG, Skopouli FN, Tzioufas AG, Moutsopoulos HM. 1998. Hypofunction of the stress axis in Sjogren’s syndrome. J. Rheumatol. 25:1508–14 Buske-Kirschbaum A, Jobst S, Psych D, Wustman A, Kirschbaum C, Rauh W, Hellhammer D. 1997. Attenuated free cortisol response to psychosocial stress in children with atopic dermatitis. Psychosom. Med. 59:419–26 Neeck G, Crofford LJ. 2000. Neuroendocrine perturbations in fibromyalgia and chronic fatigue syndrome. Rheum. Dis. Clin. North Am. 26:989–1002 Demitrack MA, Dale JK, Straus SE, Laue L, Listwak SJ, Kruesi MJ, Chrousos GP, Gold PW. 1991. Evidence for impaired activation of the hypothalamicpituitary-adrenal axis in patients with chronic fatigue syndrome. J. Clin. Endocrinol. Metab. 73:1224–34 Crofford LJ, Pillemer SR, Kalogeras KT, Cash JM, Michelson D, Kling MA, Sternberg EM, Gold PW, Chrousos GP, Wilder RL. 1994. Hypothalamicpituitary-adrenal axis perturbations in patients with fibromyalgia. Arthritis Rheum. 37:1583–92 Tsigos C. 1999. Isolated glucocorticoid

5 Feb 2002

7:56

AR

AR152-06.tex

AR152-06.SGM

LaTeX2e(2001/05/10)

P1: GJC

NEUROENDOCRINE REGULATION OF IMMUNITY

175.

Annu. Rev. Immunol. 2002.20:125-163. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

176.

177.

178.

179.

180.

181.

182.

183.

deficiency and ACTH receptor mutation. Arch. Med. Res. 30:475–80 Tsigos C, Tsiotra P, Garibaldi LR, Stravridis JC, Chrousos GP, Raptis SA. 2000. Mutation of the ACTH receptor gene in a new family with isolated glucocorticoid deficiency. Mol. Genet. Metab. 71:646–50 Wu S-M, Stratakis CA, Chan CHY, Hallermeier KM, Bourdony CJ, Rennert OM, Chan W-Y. 1998. Genetic heterogeneity of adrenocorticotropin (ACTH) resistance syndromes: identification of a novel mutation of the ACTH receptor gene in hereditary glucocorticoid deficiency. Mol. Genet. Metab. 64:256–65 Hiroi N, Yakushiji F, Shimojo M, Watanabe S, Sugano S, Yamaguchi N, Miyachi Y. 1998. Human ACTH hypersensitivity syndrome associated with abnormalities of the ACTH receptor gene. Clin. Endorinol. 48:129–34 Burke CW. 1985. Adrenocortical insufficiency. Clin. Endocrinol. Metab. 14:947–76 Kageyama Y. 2000. A case of isolated ACTH deficiency who developed autoimmune-mediated hypothyroidism and impaired water diuresis during glucocorticoid replacement therapy. Endocr. J. 47:667–74 Pugeat MM, Chrousos GP, Nisula BC, Loriaux DL, Brandon D, Lipsett MB. 1984. Plasma cortisol transport and primate evolution. Endocrinology 115:357–61 Mingrone G, DeGaetano A, Pugeat M, Capristo E, Greco AV, Gasbarrini G. 1999. The steroid resistance of Crohn’s disease. J. Invest. Med. 47:319–25 Rask E, Olsson T, Soderberg S, Andrew R, Livingstone DE, Johnson O, Walker BR. 2001. Tissue-specific dysregulation of cortisol metabolism in human obesity. J. Clin. Endocrinol. Metab. 86:1418–21 Takahashi KI, Fukushima K, Sasano H, Sasaki I, Matsuno S, Krozowski ZS, Nagura H. 1999. Type II 11beta-

184.

185.

186.

187.

188.

189.

161

hydroxysteroid dehydrogenase expression in human colonic epithelial cells of inflammatory bowel disease. Dig. Dis. Sci. 44:2516–22 Huizenga NATM, Koper JW, de Lange P, Pols HAP, Stolk RP, Burger H, Grobbee DE, Brinkmann AO, de Jong FH, Lamberts SWJ. 1998. A polymorphism in the glucocorticoid receptor gene may be associated with an increased sensitivity to glucocorticoids in vivo. J. Clin. Endocrinol. Metab. 83:144–51 Koper JW, Stolk RP, de Lange P, Huizenga NATM, Molijn G-J, Pols HAP, Grobbee DE, Karl M, de Jong FH, Brinkmann AO, Lamberts SWJ. 1997. Lack of association between five polymorphisms in the human glucocorticoid receptor gene and glucocorticoid resistance. Hum. Genet. 99:663–68 Huizenga NATM, de Lange P, Koper JW, de Herder WW, Abs R, Kasteren JHLMV, de Jong FH, Lamberts SWJ. 2000. Five patients with biochemical and/or clinical generalized glucocorticoid resistance without alterations in the glucocorticoid receptor gene. J. Clin. Endocrinol. Metab. 85:2076–81 Webster JC, Jewell CM, Bodwells JE, Munck A, Sar M, Cidlowski JA. 1997. Mouse glucocorticoid receptor phosphorylation status influences multiple functions of the receptor protein. J. Biol. Chem. 272:9287–93 Hamid QA, Wenzel SE, Hauk PJ, Tsicopoulos A, Wallaert B, Lafitte J-J, Chrousos GP, Szefler SJ, Leung DYM. 1999. Increased glucocorticoid receptor β in airway cells of glucocorticoidinsensitive asthma. Am. J. Resp. Crit. Care Med. 159:1600–4 Leung DYM, Hamid Q, Vottero A, Szefler SJ, Surs W, Minshall E, Chrousos GP, Klemm DJ. 1997. Association of glucocorticoid insensitivity with increased expression of glucocorticoid receptor β. J. Exp. Med. 186:1567–74

5 Feb 2002

7:56

Annu. Rev. Immunol. 2002.20:125-163. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

162

AR

AR152-06.tex

AR152-06.SGM

LaTeX2e(2001/05/10)

P1: GJC

WEBSTER ET AL.

190. Webster JC, Oakley RH, Jewell CM, Cidlowski JA. 2001. Proinflammatory cytokines regulate human glucocorticoid receptor gene expression and lead to the accumulation of the dominant negative β isoform: a mechanism for the generation of glucocorticoid resistance. Proc. Natl. Acad. Sci. USA 98:6865–70 191. Honda M, Orii F, Ayabe T, Imai S, Ashida T, Obara T, Kohgo Y. 2000. Expression of glucocorticoid receptor β in lymphocytes of patients with glucocorticoid-resistant ulcerative colitis. Gastroenterology 118:859–66 192. Shahidi H, Vottero A, Stratakis C, Taymans SE, Karl M, Longui CA, Chrousos GP, Daughaday WH, Gregory SA, Plate JMD. 1999. Imbalanced expression of the glucocorticoid receptor isoforms in cultured lymphocytes from a patient with systemic glucocorticoid resistance and chronic lymphocytic leukemia. Biochem. Biophys. Res. Commun. 254:559–65 193. DeRijk R, Schaaf MJM, Turner G, Datson NA, Vreugdenhil E, Cidlowski J, de Kloet ER, Emery P, Sternberg EM, Detera-Wadleigh SD. 2001. A human glucocorticoid receptor gene variant that increases the stability of the glucocorticoid receptor β-isoform mRNA is associated with rheumatoid arthritis. J. Rheumatol. In press 194. Kino T, Gragerov A, Kopp JB, Stauber RH, Pavlakis GN, Chrousos GP. 1999. The HIV-1 virion-associated protein Vpr is a coactivator of the human glucocorticoid receptor. J. Exp. Med. 189:51–61 195. New MI, Nimkarn S, Brandon DD, Cunningham-Rundles S, Wilson RC, Newfield RS, Vandermeulen J, Barron N, Russo C, Loriaux DL, O’Malley B. 1999. Resistance to several steroids in two sisters. J. Clin. Endocrinol. Metab. 84:4454–64 196. Kralli A, Yamamoto KR. 1996. An FK506-sensitive transporter selectively decreases intracellular levels and po-

197.

198.

199.

200.

201.

202.

203.

204.

205.

tency of steroid hormones. J. Biol. Chem. 271:17152–56 Medh RD, Lay RH, Schmidt TJ. 1998. Agonist-specific modulation of glucocorticoid receptor-mediated transcription by immunosuppressants. Mol. Cell. Endocrinol. 138:11–23 Meijer OC, de Lange EC, Breimer DD, de Boer AG, Workel JO, de Kloet ER. 1998. Penetration of dexamethasone into brain glucocorticoid targets is enhanced in Mdr1A P-glycoprotein knockout mice. Endocrinology 139:1789–93 Schinkel AH, Wagenaar E, van Deemter L, Mol CA, Borst P. 1995. Absence of the Mdr1a P-glycoprotein in mice affects tissue distribution and pharmacokinetics of dexamethasone, digoxin, and cyclosporin A. J. Clin. Invest. 96:1698– 705 Gottesman MM, Pastan I. 1993. Biochemistry of multidrug resistance mediated by the multidrug transporter. Annu. Rev. Biochem. 62:385–427 Farrell RJ, Murphy A, Long A, Donnelly S, Cherikuri A, O’Toole D, Mahmud N, Keeling PW, Weir DG, Kelleher D. 2000. High multidrug resistance (P-glycoprotein 170) expression in inflammatory bowel disease patients who fail medical therapy. Gastroenterology 118:279–88 Geick A, Eichelbaum M, Burk O. 2001. Nuclear receptor response elements mediate induction of intestinal MDR1 by rifampicin. J. Biol. Chem. 276:14581–87 Moore JT, Kliewer S. 2000. Use of the nuclear receptor PXR to predict drug interactions. Toxicology 153:1–10 Madden KS, Sanders VM, Felten DL. 1995. Catecholamine influences and sympathetic neural modulation of immune responsiveness. Annu. Rev. Pharmacol. Toxicol. 35:417–48 Felten DL, Felten SY. 1988. Sympathetic noradrenergic innervation of immune organs. Brain Behav. Immun. 2:293– 300

5 Feb 2002

7:56

AR

AR152-06.tex

AR152-06.SGM

LaTeX2e(2001/05/10)

P1: GJC

Annu. Rev. Immunol. 2002.20:125-163. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

NEUROENDOCRINE REGULATION OF IMMUNITY 206. Hori T, Katafuchi T, Take S, Shimizu N, Niijima A. 1995. The autonomic nervous system as a communication channel between the brain and the immune system. Neuroimmunomodulation 2:203– 15 207. Heijnen CJ, Kavelaars A. 1999. The importance of being receptive. J. Neuroimmunol. 100:197–202 208. Kohm AP, Sanders VM. 2000. Norepinephrine: a messenger from the brain to the immune system. Immunol. Today 21:539–42 209. Jansson L, Holmdahl R. 1998. Estrogenmediated immunosuppression in autoimmune diseases. Inflamm. Res. 47: 290–301 210. Kincade PW, Medina KL, Payne KJ, Rossi MI, Tudor KS, Yamashita Y, Kouro T. 2000. Early B-lymphocyte precursors and their regulation by sex steroids. Immunol. Rev. 175:128–37 211. Medina KL, Strasser A, Kincade PW. 2000. Estrogen influences the differentiation, proliferation, and survival of early

212.

213.

214.

215.

163

B-lineage precursors. Blood 95:2059– 67 Jensen EV, Greene GL, Closs LE, DeSombre ER, Nadji M. 1982. Receptors reconsidered: a 20-year perspective. In Recent Progress in Hormone Research, ed. RO Greep, pp. 1–34. New York: Academic Kuiper GG, Enmark E, Pelto-Huikko ˚ 1996. M, Nilsson S, Gustafsson J-A. Cloning of a novel receptor expressed in rat prostate and ovary. Proc. Natl. Acad. Sci. USA 93:5925–30 Erlandsson MC, Ohlsson C, Gustafsson ˚ Carlsten H. 2001. Role of oestroJ-A, gens receptors alpha and beta in immune organ development and in oestrogenmediated effects on thymus. Immunology 103:17–25 Misiewicz B, Griebler C, Gomez M, Raybourne R, Zelazowski E, Gold PW, Sternberg EM. 1996. The estrogen antagonist tamoxifen inhibits carrageenan induced inflammation in LEW/N female rats. Life Sci. 58:281–86

P1: FRK

February 1, 2002

12:47

Annual Reviews

AR152-FM

Annual Review of Immunology Volume 20, 2002

Annu. Rev. Immunol. 2002.20:125-163. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

CONTENTS FRONTISPIECE, Charles A. Janeway, Jr. A TRIP THROUGH MY LIFE WITH AN IMMUNOLOGICAL THEME, Charles A. Janeway, Jr.

xiv 1

THE B7 FAMILY OF LIGANDS AND ITS RECEPTORS: NEW PATHWAYS FOR COSTIMULATION AND INHIBITION OF IMMUNE RESPONSES, Beatriz M. Carreno and Mary Collins

29

MAP KINASES IN THE IMMUNE RESPONSE, Chen Dong, Roger J. Davis, and Richard A. Flavell

PROSPECTS FOR VACCINE PROTECTION AGAINST HIV-1 INFECTION AND AIDS, Norman L. Letvin, Dan H. Barouch, and David C. Montefiori T CELL RESPONSE IN EXPERIMENTAL AUTOIMMUNE ENCEPHALOMYELITIS (EAE): ROLE OF SELF AND CROSS-REACTIVE ANTIGENS IN SHAPING, TUNING, AND REGULATING THE AUTOPATHOGENIC T CELL REPERTOIRE, Vijay K. Kuchroo, Ana C. Anderson, Hanspeter Waldner, Markus Munder, Estelle Bettelli, and Lindsay B. Nicholson

55 73

101

NEUROENDOCRINE REGULATION OF IMMUNITY, Jeanette I. Webster, Leonardo Tonelli, and Esther M. Sternberg

125

MOLECULAR MECHANISM OF CLASS SWITCH RECOMBINATION: LINKAGE WITH SOMATIC HYPERMUTATION, Tasuku Honjo, Kazuo Kinoshita, and Masamichi Muramatsu

165

INNATE IMMUNE RECOGNITION, Charles A. Janeway, Jr. and Ruslan Medzhitov

KIR: DIVERSE, RAPIDLY EVOLVING RECEPTORS OF INNATE AND ADAPTIVE IMMUNITY, Carlos Vilches and Peter Parham ORIGINS AND FUNCTIONS OF B-1 CELLS WITH NOTES ON THE ROLE OF CD5, Robert Berland and Henry H. Wortis E PROTEIN FUNCTION IN LYMPHOCYTE DEVELOPMENT, Melanie W. Quong, William J. Romanow, and Cornelis Murre

197 217 253 301

LYMPHOCYTE-MEDIATED CYTOTOXICITY, John H. Russell and Timothy J. Ley

SIGNAL TRANSDUCTION MEDIATED BY THE T CELL ANTIGEN x

323

P1: FRK

Annu. Rev. Immunol. 2002.20:125-163. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

February 1, 2002

12:47

Annual Reviews

AR152-FM

CONTENTS

xi

RECEPTOR: THE ROLE OF ADAPTER PROTEINS, Lawrence E. Samelson INTERACTION OF HEAT SHOCK PROTEINS WITH PEPTIDES AND ANTIGEN PRESENTING CELLS: CHAPERONING OF THE INNATE AND ADAPTIVE IMMUNE RESPONSES, Pramod Srivastava CHROMATIN STRUCTURE AND GENE REGULATION IN THE IMMUNE SYSTEM, Stephen T. Smale and Amanda G. Fisher PRODUCING NATURE’S GENE-CHIPS: THE GENERATION OF PEPTIDES FOR DISPLAY BY MHC CLASS I MOLECULES, Nilabh Shastri, Susan

371

Schwab, and Thomas Serwold

395 427

463

THE IMMUNOLOGY OF MUCOSAL MODELS OF INFLAMMATION, Warren Strober, Ivan J. Fuss, and Richard S. Blumberg T CELL MEMORY, Jonathan Sprent and Charles D. Surh

GENETIC DISSECTION OF IMMUNITY TO MYCOBACTERIA: THE HUMAN MODEL, Jean-Laurent Casanova and Laurent Abel ANTIGEN PRESENTATION AND T CELL STIMULATION BY DENDRITIC CELLS, Pierre Guermonprez, Jenny Valladeau, Laurence Zitvogel, Clotilde Th´ery, and Sebastian Amigorena

495 551 581

621

NEGATIVE REGULATION OF IMMUNORECEPTOR SIGNALING, Andr´e Veillette, Sylvain Latour, and Dominique Davidson

669

CPG MOTIFS IN BACTERIAL DNA AND THEIR IMMUNE EFFECTS, Arthur M. Krieg

PROTEIN KINASE Cθ

709 IN

T CELL ACTIVATION, Noah Isakov and Amnon

Altman

RANK-L AND RANK: T CELLS, BONE LOSS, AND MAMMALIAN EVOLUTION, Lars E. Theill, William J. Boyle, and Josef M. Penninger PHAGOCYTOSIS OF MICROBES: COMPLEXITY IN ACTION, David M. Underhill and Adrian Ozinsky

761 795 825

STRUCTURE AND FUNCTION OF NATURAL KILLER CELL RECEPTORS: MULTIPLE MOLECULAR SOLUTIONS TO SELF, NONSELF DISCRIMINATION, Kannan Natarajan, Nazzareno Dimasi, Jian Wang, Roy A. Mariuzza, and David H. Margulies

853

INDEXES Subject Index Cumulative Index of Contributing Authors, Volumes 1–20 Cumulative Index of Chapter Titles, Volumes 1–20

ERRATA An online log of corrections to Annual Review of Immunology chapters (if any, 1997 to the present) may be found at http://immunol.annualreviews.org/

887 915 925

8 Feb 2002

9:40

AR

AR152-07.tex

AR152-07.SGM

LaTeX2e(2001/05/10) P1: GJC 10.1146/annurev.immunol.20.090501.112049

Annu. Rev. Immunol. 2002. 20:165–96 DOI: 10.1146/annurev.immunol.20.090501.112049 c 2002 by Annual Reviews. All rights reserved Copyright °

MOLECULAR MECHANISM OF CLASS SWITCH RECOMBINATION: Linkage Annu. Rev. Immunol. 2002.20:165-196. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

with Somatic Hypermutation Tasuku Honjo, Kazuo Kinoshita, and Masamichi Muramatsu Department of Medical Chemistry, Graduate School of Medicine, Kyoto University, Yoshida, Sakyo-ku, Kyoto 606-8501, Japan; e-mail: [email protected]; [email protected]; [email protected]

Key Words AID, RNA editing, transcription, stem-loop structure, nick endonuclease ■ Abstract Class switch recombination (CSR) and somatic hypermutation (SHM) have been considered to be mediated by different molecular mechanisms because both target DNAs and DNA modification products are quite distinct. However, involvement of activation-induced cytidine deaminase (AID) in both CSR and SHM has revealed that the two genetic alteration mechanisms are surprisingly similar. Accumulating data led us to propose the following scenario: AID is likely to be an RNA editing enzyme that modifies an unknown pre-mRNA to generate mRNA encoding a nicking endonuclease specific to the stem-loop structure. Transcription of the S and V regions, which contain palindromic sequences, leads to transient denaturation, forming the stem-loop structure that is cleaved by the AID-regulated endonuclease. Cleaved single-strand tails will be processed by error-prone DNA polymerase-mediated gap-filling or exonucleasemediated resection. Mismatched bases will be corrected or fixed by mismatch repair enzymes. CSR ends are then ligated by the NHEJ system while SHM nicks are repaired by another ligation system.

INTRODUCTION One of the most striking facts of the complete human genome sequencing is that the human genome may contain as few as 30,000 genes, only twofold more than those in the fruit fly or worm genomes (1, 2). Such a small number of genes may not be able to support highly sophisticated biological functions such as the nervous and immune systems in humans. One strategy to overcome this problem is somatic alteration of genetic information after birth. This implies that the genome is not a fixed blueprint but rather a scenario of life that requires ad libs. The immune system is known for taking advantage of a series of genetic alterations during lymphocyte differentiation. Antigen receptor genes are assembled by 0732-0582/02/0407-0165$14.00

165

1 Feb 2002

10:12

Annu. Rev. Immunol. 2002.20:165-196. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

166

AR

HONJO

AR152-07.tex

¥

AR152-07.SGM

KINOSHITA

¥

LaTeX2e(2001/05/10)

P1: GJC

MURAMATSU

site-specific recombination of subexon segments of the variable (V) region gene, namely V, diversity (D), and joining segments (J). Each step of VDJ recombination is programmed, ordered, and tightly regulated by a number of factors including cytokines provided by stromal cells. The regulation prevents generation of more than one copy of functional V exon in the heavy (H) and light (L) chain gene loci (allelic exclusion). Mature B lymphocytes, which have completed functional VDJ recombination of both H and L chain genes, express IgM on the surface and migrate to the secondary lymphoid organs such as spleen and lymph nodes where they encounter antigens. B lymphocytes activated by antigen stimulation proliferate vigorously in lymphoid follicles and often form special microenvironments called germinal centers, where the second wave of genetic alterations, namely class switch recombination (CSR) and somatic hypermutation (SHM), takes place in the immunoglobulin gene loci. CSR replaces the immunoglobulin CH gene to be expressed from Cµ to Cγ , Cε or Cα, resulting in switching of immunoglobulin isotype from IgM to either IgG, IgE, or IgA, respectively, without changing the antigen specificity. Each isotype determines the manner in which captured antigens are eliminated or the location where the immunoglobulin is delivered and accumulated. SHM takes place in the V region of both H and L chain genes, introducing a million times more point mutations than the genome-wide background. SHM followed by selection leads to generation of high-affinity antibodies. Thus, CSR and SHM generate quite distinct products in entirely different targets, i.e., CH and VL/H, respectively. Therefore, until very recently CSR and SHM were believed to be regulated by different mechanisms. However, a putative RNA editing enzyme, activation-induced cytidine deaminase (AID), has been shown to regulate both CSR and SHM in mouse and human (3, 4). The regulation of two different types of genetic alteration mechanism by AID indicates that mammals are equipped with surprisingly sophisticated and complex layers of the genetic alteration mechanisms to diversify our genomic information. This review focuses on the development of our understanding of molecular mechanisms of CSR and SHM during the past five years. We first compare basic knowledge about the molecular mechanisms of CSR and SHM and then propose a unified model that explains how AID can regulate CSR and SHM.

MICROENVIRONMENTS FOR CSR AND SHM It is generally believed that CSR and SHM take place in germinal centers (5–8). However, these two events are independent, and neither of them is a prerequisite of the other (9, 10). Cytokines provided by helper T cells and accessory cells in germinal centers regulate isotype specificity of CSR (11). Although several factors are proposed to regulate SHM in vitro (12–15), factors that regulate SHM in vivo are not clear. SHM generates a large variety of B lymphocytes expressing Ig with different affinities to a given antigen. It is therefore important to select out

1 Feb 2002

10:12

AR

AR152-07.tex

AR152-07.SGM

LaTeX2e(2001/05/10)

P1: GJC

Annu. Rev. Immunol. 2002.20:165-196. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

MECHANISM OF SWITCH RECOMBINATION

167

B lymphocytes expressing high-affinity Ig. The selection is believed to be mediated by competition for limited amounts of antigen on the surface of follicular dendritic cells located in germinal centers (16, 17). Selected B cells finally differentiate into either plasma cells or memory B cells. Several transcription factors are shown to be involved in the differentiation into plasma cells but not in CSR (18–20). Although plasma cells were believed to have no Ig on surface (from studies on cultured plasma cell lines), a recent study has clearly shown that plasma cells in vivo express Ig on surface and consist of B220+ Ig+ plasma blasts and B220− Ig+ plasma cells (21). The definition of memory B cells is more difficult. There are several surface markers, but at this stage a definitive molecular marker on memory B cells is missing, and the definition depends on its functional property. The evidence to support the notion that the germinal center is important for CSR and SHM is further obtained from studies on the mutant mice that cannot form proper lymphoid follicles and germinal centers. Aly homozygotes have neither lymph nodes nor Peyer’s patches but do have spleen with defective architecture (22). Aly mutation is caused by a point mutation in the protein-interacting domain of NFκB-inducing kinase (NIK), which abolishes the activity of lymphotoxin β signaling (23). In this mutant mouse CSR and SHM are severely damaged (24). Similarly, lymphotoxin α- and TNFα-deficient mice have a defect in the lymphoid follicular structure, although in a slightly different manner from each other; these mice have severe defects not only in the formation of germinal centers but also in CSR and SHM (25–29). Loss-of-function mutations of molecules essential for B cell activation including CD40 and its ligand (30, 31), NFκB/p52 (32, 33), c-rel (34, 35), and Bcl-3 (36) all cause defects in germinal center formation as well as in CSR. Although germinal center formation is important for CSR and SHM, it is not absolutely required. There are a number of reports that CSR takes place in the extrafollicular region (37). It is well known that IgG3 is formed in the marginal zone by T-independent antigen stimulation (38, 39). In intestine, at least a significant fraction of IgA is produced in a T-independent manner (40, 41), which suggests that IgA production in gut may depend on a unique microenvironment. Peritoneal cavity B1 cells express mature IgA mRNA (42), which suggests that B cells may switch in the peritoneal cavity, or precommitted B1 cells may arrive at the peritoneal cavity. Peritoneal cavity B1 cells have also been demonstrated to migrate to the lamina propria of the gut (43–45). However, all these findings have not shown that CSR actually occurs in situ, nor have they excluded the possibility of migration of switched cells. It is, therefore, essential to obtain a good molecular marker for ongoing CSR (37). Recently, a definitive marker for CSR was identified (37a). CSR is accompanied by excision of looped-out circular DNA. Specific transcripts are transcribed from this circular DNA, and the transcripts decay quickly in the absence of switch stimulation. Specific RT-PCR for detection of such transcripts, termed circle transcripts, can provide a hallmark for ongoing CSR in vitro and in vivo. Using this marker, a recent report clearly demonstrates that CSR from IgM to IgA takes place in situ

1 Feb 2002

10:12

Annu. Rev. Immunol. 2002.20:165-196. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

168

AR

HONJO

AR152-07.tex

¥

AR152-07.SGM

KINOSHITA

¥

LaTeX2e(2001/05/10)

P1: GJC

MURAMATSU

in the lamina propria of the intestine (46). Stromal cells derived from the lamina propria can facilitate preferential differentiation to IgA plasma cells from IgM B cells not only of the lamina propria but also of the spleen. SHM also takes place in the absence of germinal centers. Repeated or high-dose immunization can induce CSR or SHM in aly/aly, LTα −/−, and Lyn−/− mice that are defective in germinal center formation (24, 47, 48). In addition, B cells from X-linked hyper-IgM syndrome type I patients, of which the CD40 ligand gene is mutated, also have a certain level of hypermutation, although there are no germinal centers (49). Taken together, germinal center formation strongly facilitates CSR and SHM, but other environments can also support CSR and SHM.

MOLECULAR BASIS OF CSR Region-Specific Recombination The immunoglobulin CH locus consists of an ordered array of CH genes (50), each flanked at its 50 region by a switch (S) region composed of tandem repetitive sequences with many palindromes (51). CSR takes place between two S regions, resulting in loop-out deletion of intervening DNA segments as circular DNA (52– 54). Since the Cµ gene is located at the VH proximal end of the CH gene cluster, CSR between Sµ and another S region 50 to a CH gene brings that particular CH gene adjacent to the VH exon. CSR in the S regions is preceded by transcription of the two S regions starting from the I promoter located 50 to each S region. Since mutations at splicing donor sites of the transcripts reduce CSR (55, 56), splicing of transcripts appears to be important, which gives rise to germline transcripts containing the I and CH exon sequences. Structures of S regions have common features, although their exact primary sequences are diverged (57). Each S region contains conserved G-rich pentameric sequences, which are major repeat units in the Sµ region (58). In mouse, Sγ sequences are mostly repetition of 49-bp repeats, and Sε consists of 40-bp repeating units (57). Sα region consists of 80-bp unit sequences (59). Another important feature of the S region is the presence of abundant palindromic sequences, which can form the stem-loop structure in a denatured state. Similar repetitive sequences are found in S regions of human, chicken, frog, cow, pig, camel, shrew, and rabbits (60–68). Requirement of the S region for CSR was first demonstrated by an in vitro assay system using artificial switch substrates, in which the absence of S sequences completely abolished CSR (69–75). This finding is consistent with the inability to express the structurally normal human pseudo Cγ gene without the S region (76, 77). Deletion of the major portion of the Sµ core region from mouse causes a reduced frequency of CSR (78); the IgG1 production decreases to half, but clearly significant levels of many isotypes are found in sera. The results suggest that scattered pentameric unit sequences upstream of the Sµ region may serve less efficiently as a functional S region.

1 Feb 2002

10:12

AR

AR152-07.tex

AR152-07.SGM

LaTeX2e(2001/05/10)

P1: GJC

Annu. Rev. Immunol. 2002.20:165-196. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

MECHANISM OF SWITCH RECOMBINATION

169

The product of CSR is a deleted chromosomal Ig CH locus accompanied by a loop out of circular DNA. Transchromosomal recombination is also demonstrated between a transgene and the endogenous locus in mice (79) and in some part of µ-to-α switching in normal rabbits (62). Joining of cleaved ends appears to be mediated by a nonhomologous endjoining (NHEJ) repair system (80–82), which also plays an essential role in VDJ recombination. Extensive analyses of switch recombination junctions have revealed no consensus sequences in the proximity of the breakpoints nor homologous sequences between two recombined parental sequences (57, 83, 84). These results indicate that CSR is a unique type of recombination that does not belong to either homologous or site-specific recombination but belongs rather to region-specific recombination. This is consistent with the finding that primary sequences of S regions are not important for CSR (see below), although there is a report claiming the existence of class-specific factors (85). Distribution of breakpoints is mostly within the S region, but they are also found in 50 and 30 franking regions of S regions (83, 84, 86). In fact, the breakpoints are rather confined within the intronic region of germline transcripts, namely between downstream of the Iµ exon and upstream of the Cµ exon (84). The molecular mechanism of CSR can be divided in four steps: (a) selection of target S region, (b) recognition of target sequence or structure, (c) cleavage by a putative recombinase, and (d ) repair and ligation. None of these steps has been fully understood. However, recent studies using mice genetically manipulated by transgenes and gene targeting and by switching B cell lines carrying artificial switch substrates have expanded our knowledge. Furthermore, comparison of these steps with those of SHM has revealed a striking similarity, which, together with the common requirement of AID, provides an important clue for understanding of molecular mechanisms of CSR and SHM.

In Vitro CSR Systems Using Artificial Switch Substrates A number of groups reported artificial mini-chromosomal constructs to dissect the recognition target by a putative CSR recombinase (69–75, 87, 88). These constructs have many different features. It is important to evaluate each construct to determine whether it meets basic requirements of CSR in vivo. Gene disruption studies have shown that CSR depends on transcription of the S region (89–92), splicing of transcription products (55, 56), and the presence of the S region (78). CSR is absolutely dependent on AID (3, 4). From these criteria some of the systems may not represent real CSR, and the interpretation of data from such constructs is limited and so excluded from this review. Studies on artificial switch substrates have also shown the requirement of a pair of S regions (73, 75, 93, 94), their transcription (73, 75, 93, 94), and AID (K. Kinoshita, T. Honjo, unpublished data) for CSR. The splicing requirement has been partially demonstrated in an artificial system (75). New and important messages obtained from artificial substrates are (a) the I exon and C exons are dispensable (73, 75, 93, 94), and (b) the primary sequences of S regions are not important (73).

1 Feb 2002

10:12

170

AR

HONJO

AR152-07.tex

¥

AR152-07.SGM

KINOSHITA

¥

LaTeX2e(2001/05/10)

P1: GJC

MURAMATSU

SELECTION OF TARGET S REGIONS

Annu. Rev. Immunol. 2002.20:165-196. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

Role of Cytokines Cytokine stimulation activates a specific I promoter and induces synthesis of germline transcripts containing the I and CH exon sequences. Since germline transcription almost always precedes cytokine-induced CSR, the two groups (95, 96) proposed the accessibility model—that germline transcription of the S region opens its chromatin structure, allowing a putative CSR recombinase access to a particular S region. This hypothesis is generally well supported by a number of experiments that clearly indicate a close linkage of the CSR target with the transcribed S region (89–92, 97–105). The original accessibility model postulated that CSR recombinase exists before induction of germline transcription, and chromatin accessibility is limited to initiating CSR. However, artificial constructs containing a constitutive promoter for each of two S regions introduced in the CH12F3-2 B lymphoma cell line (106) were unable to switch unless cytokine stimulation was given (73). In addition, a protein synthesis inhibitor cycloheximide blocks cytokine-induced CSR, suggesting that de novo protein synthesis is required for CSR (107). Such experiments imply that cytokine stimulation plays at least two roles: (a) induction of germline transcription associated with chromatin opening and (b) induction of de novo synthesis of CSR recombinase or its activator.

Role of Germline Transcription and Transcripts If germline transcription is required only for opening the chromatin locus of S regions, a minimal level of transcription may be sufficient, and quantitative increments of transcription would not affect the CSR efficiency. A recent experiment using an artificial switch substrate containing a tetracycline-inducible promoter in place of the I promoter clearly demonstrated that germline transcription levels quantitatively correlate with the CSR efficiency (108). Experiments using transgenic loci (92, 97, 99, 100, 103) and also artificial switch constructs (73, 75, 93, 94) showed the I promoter to be dispensable and replacable by any promoters. The I promoters are regulated by signaling of cytokine receptors and CD40 (34, 109–117). Regulatory regions of many I promoters have been extensively studied and shown to contain several binding motifs of transcription factors that are regulated by specific cytokines (118–129). The involvement of enhancers has been also extensively characterized. Two DNase I hypersensitive sites, HS1,2 and HS3a in the 30 enhancer are not required for efficient CSR (105), but the intronic enhancer is required (90, 98, 101, 130). However, switch constructs without enhancers are capable of undergoing CSR if they have strong promoters for S regions (73). Enhancers are probably required in vivo to support efficient transcription. Gene replacement studies have shown that splicing disruption of germline transcripts causes severe reduction or abolishing of CSR (55, 56), which suggests that spliced germline transcripts or loading of spliceosomes on the S region may be

1 Feb 2002

10:12

AR

AR152-07.tex

AR152-07.SGM

LaTeX2e(2001/05/10)

P1: GJC

Annu. Rev. Immunol. 2002.20:165-196. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

MECHANISM OF SWITCH RECOMBINATION

171

important for CSR. Another interpretation could be that CSR is reduced because the absence of splicing decreases transcription efficiency. However, another suggestion is that splicing or spliceosome may be important; the distribution of breakpoints is closely associated with the intronic region of germline transcripts (84). Although a few reports claim the existence of translation products encoded by germline transcripts (131, 132), the finding that gene replacement of the I exon with irrelevant sequences does not affect CSR almost completely excludes the possible role of proteins encoded by germline transcripts (92, 97, 99, 100, 103). In summary, the selection of a target S region among many S regions is mediated by transcription from the particular I promoter of that S region. Since the level of transcription is correlated with the CSR efficiency, the amount of transcription machinery loaded on the S region, the stem-loop structure of the denatured S region during transcription, or spliceosomes associated with the transcribed S region may play important roles (108, 133). These three are not mutually exclusive.

RECOGNITION OF S REGIONS BY CSR RECOMBINASE The S region primary sequence is not important to CSR because, in CH12F3-2 B cell line that specifically switches to IgA, the Sα sequence of the switch construct can be replaced with the Sγ 1 or Sε sequence without changing switching efficiency (73). Inverted orientation of the Sα region is equally efficient for CSR. In vitro artificial constructs containing S regions of various species or their derivatives demonstrate that the most important features of the S region are not repetitive sequences nor G-rich sequences but palindromic sequences (134). ATrich sequences of the Xenopus S region, but not G-rich repetitive sequences of the telomere, support CSR. Most strikingly, the multiple cloning sequence of the Bluescript plasmid, which contains many palindromic sequences, was able to replace S region functionally (134), which suggests that the palindromic nature of the S region primary sequences is most important. Since palindromic sequences are rich in the S region, the stem-loop structure can be formed transiently in S regions when they are denatured during transcription (135, 136). Such a stem-loop structure is proposed as a recognition target (61, 63, 134, 137). Mußmann et al. (61) identified CSR breakpoints in close proximity to the transition sites from a stem to a loop structure, based on the singlestranded DNA folding program. Reaben & Griffin (138) found that in vitro transcription of supercoiled plasmids containing the murine Sα sequence leads to loss of a superhelical turn. Analysis of less supercoiled plasmid showed the formation of RNA/DNA hybrid by the nascent RNA transcript. Based on this in vitro observation, they proposed that the R-loop structure could be a recognition target for CSR recombinase. A similar structure was detected by in vitro transcription in a wider range of S regions (139, 140). An overexpression experiment of germline transcripts in trans and E. coli RNase H failed to support requirement of germline transcripts and R-loop formation in vivo (108).

1 Feb 2002

10:12

172

AR

HONJO

AR152-07.tex

¥

AR152-07.SGM

KINOSHITA

¥

LaTeX2e(2001/05/10)

P1: GJC

MURAMATSU

Another type of recognition structure has been proposed by the finding that single-stranded G-rich sequences self-associate to form four-stranded structure (141). This unique structure is probably held together by Hoogsteen pairing. Since the G-rich sequences exist not only in the immunoglobulin S region but also in gene promoters and chromosomal telomeres, such four-stranded structures were proposed to be recognition targets in various biological systems including CSR.

Annu. Rev. Immunol. 2002.20:165-196. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

CLEAVAGE OF THE S REGION There is no question about involvement of two double-strand cleavages in CSR, each in the S region. An important question is how they are generated. One possibility is double-strand cleavages by an endonuclease. Another possibility is two successive nicks in each S region, generating staggered double-strand cleavages. The single-strand tail of the staggered cleavage product may be processed for the subsequent joining mechanism mediated by the NHEJ repair system. The other possibility is a nick cleavage followed by transesterification as catalyzed by RAGs, producing a double-strand cleavage with a blunt end and a hairpin end (142). Wuerffel et al. (143) directly addressed this question. They purified intact genomic DNA from murine spleen B cells stimulated with LPS and examined the existence of double-strand breakage (DSB) in the Sγ 3 region by linker ligationmediated PCR. If genomic DNA contains DSB, the linker can be ligated to the cleaved end, and subsequent PCR with the primer pair, one on the linker and the other on the Sγ 3 flank, can amplify specific DNA fragments. They could amplify Sγ 3 fragments from DNA of LPS-stimulated B cells and determine the sequence of one B cell–specific band. Although similar sequences are identified in other parts of the Sγ 3 region, only the one identified by ligation-mediated PCR is surrounded by the palindromic sequence that can form the stem-loop structure. Therefore, they speculate that the secondary structure could be a recognition target for cleavage. This cleavage site is located within the SNAP region that they previously reported as a frequent recombination junction sequence and a biding site of a protein complex purified from switching B cells (144). Based on such in vivo cleavage site analysis, SNIP/SNAP sequences are proposed to be a recognition target of the recombinase (144–146). The experiments, however, have several limitations that prevent concluding that the blunt end cleavage is the primary product of CSR recombinase. First, linker-mediated PCR can detect not only blunt ends but also nicked cleavage ends. Second, it is possible that blunt-ended DSB are generated after processing of staggered ends. Third, it is not known whether blunt-ended DSB are intermediates of CSR or accidental cleavage products by activation of B cells. Previous sequence analyses of breakpoints in looped-out circles and deleted chromosomal loci did not tell whether there is any deletion or duplication after recombination because it is impossible to find a correct pair of products. If the two breakpoints of a single recombination event are retained in the substrate by inversion-type recombination, one can estimate the mode of cleavage by

8 Feb 2002

10:27

AR

AR152-07.tex

AR152-07.SGM

LaTeX2e(2001/05/10)

P1: GJC

Annu. Rev. Immunol. 2002.20:165-196. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

MECHANISM OF SWITCH RECOMBINATION

173

examination of deletion or duplication during CSR. As shown for VDJ recombination, CSR can theoretically generate inversion products. In fact, the organization of the chicken CH locus suggests inversion-type CSR in vivo (147). Kinoshita et al. (148) have shown that inversion-type CSR is enhanced in switch substrates when two S regions are transcribed by two separate promoters in the opposite direction. Using such substrates, nucleotide sequences of junction points in more than 30 inversion-type switch products were determined (148a). The majority of them contain either duplication or deletion. As shown in Figure 1, duplication during recombination can be explained only by staggered cleavage, followed by DNA synthesis to convert single-stranded tails to double-stranded ends. Deletion of sequences can be explained by multiple blunt end cleavages or by a staggered cleavage, followed by exonuclease chewing of single-stranded tails. It is known that cleaved ends of the CSR target have to be double-stranded for the NHEJ repair system. The distance of two nicks can vary, and variable lengths of deletion or duplication can be explained by variation in the distance of two nicks. The results suggest that staggered cleavage is more likely than blunt end double-strand cleavage and nick coupled with transesterification as the first step of CSR cleavage.

REPAIR AND JOINING OF CLEAVED ENDS The cleaved ends of S regions have to be repaired and joined together to give rise to looped-out circular DNA and deleted chromosome. The NHEJ system clearly is involved in joining of cleaved ends of S regions. SCID mice contain a leaky mutation in DNA-PKcs. SCID-derived pre-B cell lines cultured on stromal cells are incapable of switching (80). On the other hand, wild-type and RAG-2 mutant cells can undergo CSR. Since SCID pre-B cell lines can induce germline transcripts, they do not have a defect in activation of the locus. Ku-80 and Ku-70, which form a complex with DNA-PKcs to function as DNA-PK, are shown to be required for CSR (81, 82). RAD51, which is involved in homologous recombination, is induced in LPSstimulated B cells (149–151). RAD 51 induction is restricted to germinal center B cells, suggesting some roles of its protein in either SHM, CSR, or prevention of cell death due to excessive or aberrant cleavage. RAD54 knockout experiments demonstrate that RAD54 is not required for VDJ recombination or CSR (152). However, this does not exclude involvement of homologous recombination in CSR because in the absence of RAD54, homologous recombination is mediated by other components in the RAD52 pathway (152). The single-stranded tail of the staggered cleavage has to be either chewed by exonucleases or gap-filled by DNA synthesis to be ligated by the NHEJ repair system. Although involvement of exonuclease activity specifically induced in B cells (153, 154) is postulated, its biological function still remains to be determined. Involvement of error-prone DNA synthesis is supported by frequent mutations in

1 Feb 2002

10:12

Annu. Rev. Immunol. 2002.20:165-196. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

174

AR

HONJO

AR152-07.tex

¥

AR152-07.SGM

KINOSHITA

¥

LaTeX2e(2001/05/10)

P1: GJC

MURAMATSU

the proximity of CSR breakpoints (86, 155, 156). Most recently a number of errorprone DNA polymerases are identified in mammals (157). These polymerases may play important roles in repair of cleaved ends in CSR. Extensive search for possible involvement of repair enzymes has been done. Involvement of nucleotide excision repair systems including XP-B, D, V, C, and CS-A in CSR was examined using cell lines isolated from patients with a defect in nucleotide excision repair. Epstein-Barr virus–transformed B cell lines from these patients showed no defect in either SHM or CSR, excluding involvement of nucleotide excision repair enzymes (158, 159). Involvement of mismatch repair enzymes in CSR was also extensively studied. Msh2-deficient animals show reduction in IgG switching and deviation of CSR breakpoints (160). Studies on various mismatch repair–deficient animals, including Msh2, Mlh1/ Pms2 or Mlh1/ Pms2 mutants, have shown reduction of CSR to variable extents ranging between 35% and 75% (161). Interestingly, none of the defects are absolute, suggesting that these enzymes may play some roles in repairing ends, but many enzymes can complement each other. There may be some sequence preference of mismatch repair enzymes, which leads to the deviation of breakpoints in Msh2-deficient animals (160).

PROTEIN CANDIDATES PROPOSED TO BE INVOLVED IN CSR A number of groups tried to purify proteins that can specifically bind to S regions. Sµbp-2 was isolated as a single-strand DNA-binding protein recognizing the Sµ motif (162). The structure of Sµbp-2 indicates that it belongs to the helicase superfamily involved in replication, recombination, and repair. Recently, Sµbp-2 was shown to be involved in regulation of various genes (163–166). Binding proteins to Sγ 1 and Sα are partially purified (167, 168). SNAP, which was originally identified as an Sγ 3 bind protein (144), contains at least two subunits. One of the components shares an epitope with E47 transcription factor (169). Several other known transcription factors bind S regions (144, 170, 171). A transcription factor E2A was shown to be involved in CSR, using dominant negative experiments (172, 173), although the precise step of its involvement is not clear. NFκB binds a specific region of Sγ 3 (144). LR1 induced in LPS-stimulated B cells binds S regions (174). Highly phosphorylated LR1 contains 106-kDa nucleolin, which is involved in regulation of ribosomal DNA transcription and ribosomal RNA processing in nucleoli (175). The other component of LR1 is hnRNP D (176). LR1 binds not only Sγ 1, Sγ 3, and Sα but also Eµ enhancer (177). Several candidate enzymes are proposed for cleaving S regions. Topoisomerase II is a nicking enzyme but normally does not cleave the S region. When G-rich tetraplex is formed in the S region, topoisomerase II can cut such a structure (178). Since excision repair nucleases XPF-ERCC1 and XPG can cleave the R-loops formed in the S region during in vitro transcription, these nucleases are proposed to be involved in cleaving S regions (140). By in vitro recombination assay using

1 Feb 2002

10:12

AR

AR152-07.tex

AR152-07.SGM

LaTeX2e(2001/05/10)

P1: GJC

MECHANISM OF SWITCH RECOMBINATION

175

Annu. Rev. Immunol. 2002.20:165-196. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

plasmids containing S region, a putative recombinase complex was isolated (179). This complex contains several components such as nucleolin, poly(ADP-ribose)polymerase, nucleophosmin, and SWAP70. SWAP70 has lately been shown to be present in the cytoplasm and to form a complex with surface Ig, although its function is unknown (180, 181). SWAP70 is augmented by stimulation and is abundant in germinal center B cells. Purification of S region–specific endonuclease (Endo-SR) is reported (182, 183). All these interesting candidate proteins remain to be further investigated to determine whether they play any essential role in CSR.

SOMATIC HYPERMUTATION Distribution of Mutations and Their Target Specificity Lines of evidence have clearly shown that the complementarity determining regions (CDRs) are preferred targets of mutation to the framework regions of the Ig V regions, as originally proposed by Wu et al. (184). The three-dimensional structure of Ig has established the functional significance of CDR as an antigen binding site. Subsequently, extensive studies have been carried out to look for any primary sequences associated with mutation sites, and these revealed a few preferred motifs, among which the RGYW motif is approximately twofold more frequently mutated than by chance (185–188). However, because of the selection after hypermutation, whether CDR and the RGYW motif are the preferred targets of the mutation event or selection has to be examined. To solve this problem, nucleotide sequences of nonfunctional V genes, especially out-of-frame VDJ recombination products, were investigated. Such studies indicate that CDRs are preferred targets of mutation, as compared with framework regions, although there is obvious selection for CDR (189). In addition, the RGYW motif contains the mutation more frequently even in nonfunctional VH genes. Furthermore, contrary to previous suggestions (190, 191), there is no strand bias for the mutation target (188, 192). The importance of primary sequences is confirmed by observing the reduction or increase of mutability after changing a few bases in transgenes (185, 193). However, the primary sequence is not the only determinant of mutation targets because not all RGYW motifs are mutated, and the RGYW motif inserted in other environments loses its mutability (185). It is interesting to note that the RGYW motif includes the AGCT palindromic sequence, which is most abundantly found in S sequences. The microsequence specificity of mutations introduced during SHM and those introduced meiotically during neutral evolution is generally similar, which suggests that the enzyme machinery incorporating mutations may be shared (194). The Ig gene is not the only target of hypermutation (195–200). The bcl-6 gene in human B cells accumulates mutations with slightly lower frequency (195, 196). Translocated c-myc genes also contain frequent mutations (199, 200). Many other genes like β-globin accumulate extensive mutations when driven by the Ig promoter and intron enhancer as transgenes (201–205). Interestingly, insertion of the

1 Feb 2002

10:12

Annu. Rev. Immunol. 2002.20:165-196. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

176

AR

HONJO

AR152-07.tex

¥

AR152-07.SGM

KINOSHITA

¥

LaTeX2e(2001/05/10)

P1: GJC

MURAMATSU

EPS sequence (tandem restriction site sequences for EcoRV and PvuII), which contains abundant palindrome sequences, strongly augments the SHM frequency in surrounding V gene sequences (201, 205). The EPS insert is mutated many times more frequently than the flanking Ig sequences. Interestingly, Kolchanov et al. (206) pointed out that not only V genes but also other hypermutated genes including c-myc contain palindromic sequences. Furthermore, CDRs overlap the stem-loop structure predicted by the computer program (206). Since Brenner & Milstein (207) proposed that hypermuation is introduced by DNA cleavage and error-prone repair, initiation of SHM by DNA cleavage has been taken for granted. It is important to note that the primary DNA cleavage site recognized by a putative endonuclease of SHM and the actual base change site detected as mutations may not be the same. We can only determine the outcome of base changes that are probably generated during repair or DNA synthesis after DNA cleavage. Taking all these results together, the following conclusion is drawn for SHM target specificity: (a) some motifs are preferred but not absolutely required, and (b) palindromic sequences are not only preferred targets but are also stimulators of SHM. Therefore, the stem-loop structure based on palindromic sequences is a most likely candidate for a recognition target of SHM endonuclease.

Quantitative Correlation of V Region Transcription with SHM Frequency The transcription requirement of hypermutation target genes has been demonstrated by several convincing experiments (202, 208–210). Transgenic mice carrying a transgene, in which the Ig promoter is duplicated upstream of the Cκ region but downstream of the intronic enhancer, accumulate SHM not only in the V gene but also in the Cκ gene, which normally does not mutate (202). The authors postulate that a putative mutator is guided to the target DNA by RNA polymerase. Another series of transgenic experiments, including deletion of V promoter and replacement of V promoter with that of RNA polymerase I, have clearly demonstrated that the level of transcription parallels the frequency of SHM (209). The most direct quantitative correlation between the transcription level and SHM frequency is demonstrated using 18-81 pre B cell transfectants with a GFP transgene containing a point mutation to block its expression. Transcription of the GFP gene is controlled by the tetracycline-inducible promoter, and the level of transcripts from the GFP gene is almost directly correlated with the frequency of SHM measured by expression of GFP (210). Methylation is responsible for suppression of SHM probably through reducing the transcriptional efficiency (211, 212). The distance from the promoter determines the initiation site of mutation in the target gene. Several trangenes that contain insertion and deletion between the promoter and V gene shifted not only distribution (202, 213) but also frequency of mutations (214, 215). Again the efficiency of transcription has to be carefully evaluated to compare the frequency of mutations in various transgenes.

1 Feb 2002

10:12

AR

AR152-07.tex

AR152-07.SGM

LaTeX2e(2001/05/10)

P1: GJC

MECHANISM OF SWITCH RECOMBINATION

177

The enhancer requirement for SHM was shown by deletion of Eµ and 30 κ enhancers from transgenes (215–217). In turn, insertion of the enhancer facilitated SHM in non-Ig genes (215, 218). However, the transcription level may be of primary importance, and the presence of the enhancer may not be crucial (219), in agreement with the fact that the tetracycline-inducible promoter of the mutating GFP gene described above does not have an enhancer (210).

Annu. Rev. Immunol. 2002.20:165-196. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

Cleavage of Target DNA Evidence to demonstrate that DNA cleavage in the V gene is an obligatory intermediate of SHM is limited in spite of general acceptance. Sale & Neuberger (220) reported that extra nucleotides are inserted preferentially into the V gene of hypermutating Ramos B cell line when terminal deoxynucleotidyl transferase is overexpressed. The distribution of inserted oligonucleotides agrees broadly with the general distribution of point mutations. Extensive deletions or insertions are also observed in V genes of human memory B cells, which probably reflect aberrant products of SHM (221, 222). More recently the ligation-mediated PCR method was employed to detect DSB in hypermutating B cells (223, 224). These studies have shown that distribution of DSB overlaps with CDR and the RGYW motif. Although these experiments show that DSBs exist in the V gene at much higher frequency than that in the C gene, PCR amplification does not allow comparison of the relative frequency of two loci when different primers are used. It is important to compare DSB in the rearranged and germline V genes. Since background DSBs are not rare in proliferating cells (225), it is important to assess which fraction of DSB are the intermediate of hypermutation. Furthermore, the sites of mutations do not necessarily correlate with those of cleavage. In case of CSR, recombination joining sites (probably cleavage sites) and mutation sites can be ∼300 bp apart (155). Whether DSB are either the primary product or secondary repair product of nicking should be also examined. Kong & Meizels (226), using a similar method, have found that the majority of incidents of cleavage in the V region of mutating B cells are singlestranded nicks.

Generation and Fixation of Mutations Incorporation of mismatched bases and subsequent fixation as mutations depend on DNA synthesis and repair, respectively. Weill and his colleagues (227) have studied whether SHM is due either to suppression of repair of mismatched bases that can be incorporated during global DNA replication or to error-prone DNA synthesis of short-range outside global DNA replication. They have generated a V gene transgene with TG tandem repeat insert as an artificial microsatellite, and they found that point mutations dissociate from frame-shift mutations that are generated by slippage during global DNA replication. They conclude that hypermutation is incorporated by error-prone DNA synthesis and not by suppression of mismatch repair enzymes, which should take care of general errors during DNA

1 Feb 2002

10:12

Annu. Rev. Immunol. 2002.20:165-196. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

178

AR

HONJO

AR152-07.tex

¥

AR152-07.SGM

KINOSHITA

¥

LaTeX2e(2001/05/10)

P1: GJC

MURAMATSU

replication. To examine the role of mismatch repair enzymes in SHM, distribution and frequency of SHM are studied in knockout animals in relation to mismatch repair enzymes such as Msh2, 3, 6, Pms2, and Mlh1 (228–237). Mutations of some of these enzymes, especially Msh2, reduced the frequency of SHM and shifted targets of hot spot toward G to C biased (232–235). The results are interpreted in three different ways. In one way, mismatch repair enzymes are actively involved to increase the frequency of SHM (233, 236). In another, mismatch repair enzymes are involved in fixation of misincorporated bases (231, 232, 235, 238). In the third, observations obtained from mismatch repair–deficient mice can be an indirect effect due to altered cellular viability, the result of general genomic instability (234, 237). However, the biased base change in mismatch repair enzyme-deficient mice could be best explained by the hypothesis that mismatch repair is involved in fixation of SHM, and there is redundancy among mismatch repair enzymes. The nucleotide excision repair system is not directly involved in SHM. Frequency of SHM is normal in Epstein-Barr virus–transformed B cells from patients defective in XP-B, D, V, and CS-A (158). Similarly mice defective in the XP-A, D, C, CS-B, poly ADP-ribose polymerase, and Rad54 genes are almost normal in SHM (159, 231, 232). The NHEJ repair system does not appear to play a role in SHM (239). A considerable list of error-prone DNA polymerases was recently identified (157). Among these DNA polymerases η and ζ are reported to be involved in SHM (240–243). Since the absence of any one of these polymerases does not completely abolish SHM, error-prone polymerases appear to be redundant (241). However, they have some preference for incorporation of bases, as has been characterized by biochemical properties of polymerase η (244–246). These results suggest a scenario that after cleavage of V genes, error-prone DNA polymerases introduce mismatched bases, which will be either fixed or corrected by mismatch repair enzymes.

THE REQUIREMENT OF AID FOR BOTH CSR AND SHM Since CSR depends on de novo protein synthesis, cDNA subtraction between switch-induced and non-induced CH12F3-2 cells was carried out to identify a class-switch recombinase and/or its activator. AID thus isolated is expressed specifically in B cells activated in vivo or in vitro. In situ hybridization showed that AID expression is restricted to germinal centers. (107) Overexpression of AID in CH12F3-2 B cells enhances CSR, irrespective of cytokine stimulation. AID-deficient mice cannot produce IgG, IgA, or IgE antibodies, whereas IgM is more abundantly produced under immunized or non-immunized conditions (3). Immunoglobulin classes other than IgM and IgD are not detected, even after LPS and cytokine stimulation of spleen cells in vitro. Sequence analysis of the VH186.2 gene in AID−/− mice after immunization with 4-hydroxy-3nitrophenylacetyl (NP) conjugated with chicken γ -globulin has revealed that the

1 Feb 2002

10:12

AR

AR152-07.tex

AR152-07.SGM

LaTeX2e(2001/05/10)

P1: GJC

Annu. Rev. Immunol. 2002.20:165-196. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

MECHANISM OF SWITCH RECOMBINATION

179

mutation frequency is no more than the error rate of the Taq polymerase used for the experiment. A mutation in the CD40L gene has been demonstrated as a cause of a condition called X-linked hyper-IgM syndrome type 1 (HIGM1), which manifests as severe immunodeficiency due to a defect in CSR (247). There is also another type of HIGM (HIGM2) that shows similar clinical symptoms but is inherited in an autosomal-recessive manner. Both the human AID gene and the HIGM2 locus have been mapped to chromosome 12p13 (248), and the AID coding region from 36 HIGM2 patients has various mutations, all of which would give rise to mutant AID proteins with amino acid replacements or truncations (4, 249). Moreover, SHM is not observed in B cells from HIGM2 patients. These findings in human and mouse convincingly demonstrate the requirement of AID for both CSR and SHM. In vitro stimulation of B cells from AID-deficient mice with LPS and cytokines induces normal levels of germline transcription of specific S regions, implying that AID is involved neither in signal transduction from the cell surface receptor to the nucleus, nor in the establishment of the accessible state of the S region. Indeed, AID-deficient mouse and human have enlarged germinal centers. V(D)J recombination is normal in humans and mice with the AID mutation, indicating that AID is not involved in the repair process shared between V(D)J recombination and CSR. In other words, AID is most likely to be involved in the cleavage step in CSR. Recently, recruitment of Nbs1 and γ -H2AX to CSR chromosal DSBs are shown to be dependent on AID (S. Petersen, R. Casselas, H. T. Chen, M. J. Difilippantonio, T. Ried, M. Muramatsu, T. Honjo, M. C. Nussenzweig, A. Nussenzweig, manuscript submitted). The predicted amino acid sequence (198 residues) of AID has homology to that of APOBEC-1, which is a catalytic subunit of the RNA-editing complex for the apolipoprotein B messenger RNA precursor (107). APOBEC-1 forms a complex with an RNA-binding protein ACF and converts the specific cytosine at position 6666 of apolipoprotein B mRNA to uracil by its intrinsic cytidine deaminase activity, giving rise to mRNA encoding the chylomicron component (250, 251). Like APOBEC-1, AID synthesized in vitro possesses cytidine deaminase activity (107). Moreover, both the APOBEC-1 and AID genes are closely mapped to human chromosome 12p13 (248). These structural, genetic, and functional relationships of AID and APOBEC-1 suggest that AID may be an RNA-editing enzyme. AID in a complex with an ACF-like protein may edit a certain mRNA, giving rise to a new mRNA-encoding class switch recombinase.

A HYPOTHESIS OF A MOLECULAR MECHANISM COMMON TO CSR AND SHM CSR and SHM share many important features although their targets and products are distinct. The most striking common features for the two processes are (a) AID requirement, (b) quantitative correlation of target transcription with

8 Feb 2002

10:29

Annu. Rev. Immunol. 2002.20:165-196. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

180

AR

HONJO

AR152-07.tex

¥

AR152-07.SGM

KINOSHITA

¥

LaTeX2e(2001/05/10)

P1: GJC

MURAMATSU

genetic alteration, and (c) association of DNA cleavage with mutations. We first postulate that AID is an RNA-editing enzyme that converts pre-mRNA encoding an unknown protein to mRNA coding for a nicking endonuclease specific to the stem-loop structure (Figure 1). As discussed above, the stem-loop structure is speculated to be the recognition target for both CSR and SHM. Because the RNA polymerase holoenzyme complex contains a DNA helicase component, the DNA strands in the proximity of actively transcribing RNA polymerase are thought to be transiently denatured. The secondary structure in the singlestranded DNA may then persist long enough to be recognized by the recombinase or mutator. The endonuclease(s) encoded by AID-edited mRNA cleaves the proximity of the stem-loop structure in the V and S regions during SHM and CSR reactions, respectively (51). Transcription of S regions may be required not only for opening the chromatin structure, but also to form the secondary structure that will be recognized by the endonuclease (Figure 1). Because such recognition and cleavage take place separately on each strand, cleavage results in either a singlestranded or double-stranded staggered nick. In the case of CSR, the majority of cleaved ends are the staggered cleavage of two strands because of the abundance of the stem-loop structure. Single-strand tails of staggered cleavage ends must be either filled or resected before ligation by the NHEJ system. This may be carried out by exonucleases and/or error-prone DNA polymerases (157), which introduce mutations near the cleaved and ligated junctions. In the case of SHM, on the other hand, the stem-loop structure in the V gene may be attacked by the same or a similar endonuclease, but less efficiently, probably owing to less of the secondary structure. This cleavage results in the generation of single-stranded nicks, which could be repaired by low-fidelity DNA polymerases, exonucleases, and ligases. Mismatch repair enzymes are most likely to be involved in fixation of mutations in CSR as well as SHM. In some cases, AIDinduced cleavage in V genes also takes place on two strands, generating doublestranded staggered ends, the single-stranded tails of which may be processed and repaired as described above for CSR. Although recent reports (223, 224) suggest the presence of double-stranded blunt ends in the V gene of the hypermutating B cells, nick and/or staggered cleavages could be primary (226, 252). The joining systems of cleaved ends for CSR and SHM are distinct (80–239). This is consistent with the assumption that the majority of cleavages in CSR and SHM are staggered double-stranded cleavages and single-stranded nicks, respectively. It has been well established that general DSBs are repaired by either the NHEJ system or homologous recombination, depending on the cell cycle (253). DSBs in the V region were proposed to be repaired by homologous recombination because they accumulate specifically in the S/G2 phase of the cell cycle (224). If this is the case the cleaving enzyme is likely to be induced only during the S/G2 phase. Expression of germline transcripts and their regulatory proteins occurs mainly in the G1/S phase (254). Since AID-dependent localization of Nbs1 and γ -H2AX at DSB in the IgH locus is most prominent in the G1 phase of the cell cycle (S. Petersen, R. Casselas, H. T. Chen, M. J. Difilippantonio, T. Ried, M.

1 Feb 2002

10:12

AR

AR152-07.tex

AR152-07.SGM

LaTeX2e(2001/05/10)

P1: GJC

MECHANISM OF SWITCH RECOMBINATION

181

Muramatsu, T. Honjo, M. C. Nussenzweig, A. Nussenzweig, manuscript submitted), AID could differentially regulate SHM and CSR. It is possible that AID edits different pre-mRNA in CSR and SHM. One advantage of this model is that it is easy to explain why SHM does not always occur simultaneously with CSR.

Annu. Rev. Immunol. 2002.20:165-196. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

CONCLUSION AND PERSPECTIVE Molecular mechanisms of CSR and SHM are compared and found to share many important features in spite of the distinct products of their genetic alteration. Recent discovery of AID involvement in CSR as well as SHM led us to propose the hypothesis that a common nicking endonuclease, generated by RNA-editing activity of AID, recognizes and cleaves the stem-loop structure formed during transcription of the S and V region DNAs. Cleaved ends will be processed by exonuclease, error-prone DNA polymerases, and mismatch repair enzymes. To test this model, it is essential to prove that AID is an RNA-editing enzyme. Then identification of the target pre-mRNA of AID should provide the critical clue for understanding the basic mechanism for both CSR and SHM. It is also important to know whether AID requires ACF-like cofactors for its function. Taken together, studies on AID function appear to be a most direct approach to elucidate the remaining mysteries in the molecular immunology, i.e., mechanisms of CSR and SHM. ACKNOWLEDGMENTS We are grateful to Drs. Y. Sakakibara, A. Shimizu, S. Fagarasan, and H. Nagaoka for their critical reading of the manuscript. We also thank K. Saito for her preparation of the manuscript. This investigation is supported by the Center of Excellence Grant from the Ministry of Education, Science, Sports, and Culture of Japan. Visit the Annual Reviews home page at www.annualreviews.org

LITERATURE CITED 1. Venter JC, Adams MD, Myers EW, Li PW, Mural RJ, Sutton GG, Smith HO, Yandell M, Evans CA, Holt RA, et al. 2001. The sequence of the human genome. Science 291:1304–51 2. Lander ES, Linton LM, Birren B, Nusbaum C, Zody MC, Baldwin J, Devon K, Dewar K, Doyle M, FitzHugh W, et al. 2001. Initial sequencing and analysis of the human genome. Nature 409:860–921 3. Muramatsu M, Kinoshita K, Fagarasan S, Yamada S, Shinkai Y, Honjo T. 2000.

Class switch recombination and hypermutation require activation-induced cytidine deaminase (AID), a potential RNA editing enzyme. Cell 102:553–63 4. Revy P, Muto T, Levy Y, Geissmann F, Plebani A, Sanal O, Catalan N, Forveille M, Dufourcq-Labelouse R, Gennery A, et al. 2000. Activation-induced cytidine deaminase (AID) deficiency causes the autosomal recessive form of the hyperIgM syndrome (HIGM2). Cell 102:565– 75

1 Feb 2002

10:12

Annu. Rev. Immunol. 2002.20:165-196. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

182

AR

HONJO

AR152-07.tex

¥

AR152-07.SGM

KINOSHITA

¥

LaTeX2e(2001/05/10)

P1: GJC

MURAMATSU

5. Berek C, Berger A, Apel M. 1991. Maturation of the immune response in germinal centers. Cell 67:1121–29 6. Pascual V, Liu YJ, Magalski A, de Bouteiller O, Banchereau J, Capra JD. 1994. Analysis of somatic mutation in five B cell subsets of human tonsil. J. Exp. Med. 180:329–39 7. Jacob J, Kelsoe G, Rajewsky K, Weiss U. 1991. Intraclonal generation of antibody mutants in germinal centres. Nature 354:389–92 8. Liu YJ, Malisan F, de Bouteiller O, Guret C, Lebecque S, Banchereau J, Mills FC, Max EE, Martinez-Valdez H. 1996. Within germinal centers, isotype switching of immunoglobulin genes occurs after the onset of somatic mutation. Immunity 4:241–50 9. Rudikoff S, Pawlita M, Pumphrey J, Heller M. 1984. Somatic diversification of immunoglobulins. Proc. Natl. Acad. Sci. USA 81:2162–66 10. Siekevitz M, Kocks C, Rajewsky K, Dildrop R. 1987. Analysis of somatic mutation and class switching in naive and memory B cells generating adoptive primary and secondary responses. Cell 48:757–70 11. Snapper CM, Marcu KB, Zelazowski P. 1997. The immunoglobulin class switch: beyond “accessibility.” Immunity 6:217– 23 12. Wu H, Pelkonen E, Knuutila S, Kaartinen M. 1995. A human follicular lymphoma B cell line hypermutates its functional immunoglobulin genes in vitro. Eur. J. Immunol. 25:3263–69 13. Bergthorsdottir S, Gallagher A, Jainandunsing S, Cockayne D, Sutton J, Leanderson T, Gray D. 2001. Signals that initiate somatic hypermutation of B cells in vitro. J. Immunol. 166:2228–34 14. Dahlenborg K, Pound JD, Gordon J, Borrebaeck CA, Carlsson R. 2000. Signals sustaining human immunoglobulin V gene hypermutation in isolated germinal centre B cells. Immunology 101:210–17

15. Zan H, Cerutti A, Dramitinos P, Schaffer A, Li Z, Casali P. 1999. Induction of Ig somatic hypermutation and class switching in a human monoclonal IgM+ IgD+ B cell line in vitro: definition of the requirements and modalities of hypermutation. J. Immunol. 162:3437–47 16. Rajewsky K. 1996. Clonal selection and learning in the antibody system. Nature 381:751–58 17. MacLennan IC. 1994. Germinal centers. Annu. Rev. Immunol. 12:117–39 18. Reimold AM, Iwakoshi NN, Manis J, Vallabhajosyula P, Szomolanyi-Tsuda E, Gravallese EM, Friend D, Grusby MJ, Alt F, Glimcher LH. 2001. Plasma cell differentiation requires the transcription factor XBP-1. Nature 412:300–7 19. Piskurich JF, Lin KI, Lin Y, Wang Y, Ting JP, Calame K. 2000. BLIMP-I mediates extinction of major histocompatibility class II transactivator expression in plasma cells. Nat. Immunol. 1:526–32 20. Mittrucker HW, Matsuyama T, Grossman A, Kundig TM, Potter J, Shahinian A, Wakeham A, Patterson B, Ohashi PS, Mak TW. 1997. Requirement for the transcription factor LSIRF/IRF4 for mature B and T lymphocyte function. Science 275:540–43 21. Kamata T, Nogaki F, Fagarasan S, Sakiyama T, Kobayashi I, Miyawaki S, Ikuta K, Muso E, Yoshida H, Sasayama S, Honjo T. 2000. Increased frequency of surface IgA-positive plasma cells in the intestinal lamina propria and decreased IgA excretion in hyper IgA (HIGA) mice, a murine model of IgA nephropathy with hyperserum IgA. J. Immunol. 165:1387– 94 22. Miyawaki S, Nakamura Y, Suzuka H, Koba M, Yasumizu R, Ikehara S, Shibata Y. 1994. A new mutation, aly, that induces a generalized lack of lymph nodes accompanied by immunodeficiency in mice. Eur. J. Immunol. 24:429–34 23. Shinkura R, Kitada K, Matsuda F, Tashiro K, Ikuta K, Suzuki M, Kogishi K,

1 Feb 2002

10:12

AR

AR152-07.tex

AR152-07.SGM

LaTeX2e(2001/05/10)

P1: GJC

MECHANISM OF SWITCH RECOMBINATION

Annu. Rev. Immunol. 2002.20:165-196. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

24.

25.

26.

27.

28.

29.

30.

Serikawa T, Honjo T. 1999. Alymphoplasia is caused by a point mutation in the mouse gene encoding NF-κB-inducing kinase. Nat. Genet. 22:74–77 Shinkura R, Matsuda F, Sakiyama T, Tsubata T, Hiai H, Paumen M, Miyawaki S, Honjo T. 1996. Defects of somatic hypermutation and class switching in alymphoplasia (aly) mutant mice. Int. Immunol. 8:1067–75 Ryffel B, Di Padova F, Schreier MH, Le Hir M, Eugster HP, Quesniaux VF. 1997. Lack of type 2 T cell-independent B cell responses and defect in isotype switching in TNF-lymphotoxin α-deficient mice. J. Immunol. 158:2126–33 Pasparakis M, Alexopoulou L, Episkopou V, Kollias G. 1996. Immune and inflammatory responses in TNFα-deficient mice: a critical requirement for TNFα in the formation of primary B cell follicles, follicular dendritic cell networks and germinal centers, and in the maturation of the humoral immune response. J. Exp. Med. 184:1397–411 Fu YX, Molina H, Matsumoto M, Huang G, Min J, Chaplin DD. 1997. Lymphotoxin-α (LTα) supports development of splenic follicular structure that is required for IgG responses. J. Exp. Med. 185:2111– 20 Le Hir M, Bluethmann H, Kosco-Vilbois MH, Muller M, di Padova F, Moore M, Ryffel B, Eugster HP. 1996. Differentiation of follicular dendritic cells and full antibody responses require tumor necrosis factor receptor-1 signaling. J. Exp. Med. 183:2367–72 Eugster HP, Muller M, Karrer U, Car BD, Schnyder B, Eng VM, Woerly G, Le Hir M, di Padova F, Aguet M, et al. 1996. Multiple immune abnormalities in tumor necrosis factor and lymphotoxin-α double-deficient mice. Int. Immunol. 8: 23–36 Xu J, Foy TM, Laman JD, Elliott EA, Dunn JJ, Waldschmidt TJ, Elsemore J, Noelle RJ, Flavell RA. 1994. Mice defi-

31.

32.

33.

34.

35.

36.

37.

183

cient for the CD40 ligand. Immunity 1:423–31 Kawabe T, Naka T, Yoshida K, Tanaka T, Fujiwara H, Suematsu S, Yoshida N, Kishimoto T, Kikutani H. 1994. The immune responses in CD40–deficient mice: impaired immunoglobulin class switching and germinal center formation. Immunity 1:167–78 Caamano JH, Rizzo CA, Durham SK, Barton DS, Raventos Suarez C, Snapper CM, Bravo R. 1998. Nuclear factor (NF)-κB2 (p100/p52) is required for normal splenic microarchitecture and B cellmediated immune responses. J. Exp. Med. 187:185–96 Franzoso G, Carlson L, Poljak L, Shores EW, Epstein S, Leonardi A, Grinberg A, Tran T, Scharton-Kersten T, Anver M, et al. 1998. Mice deficient in nuclear factor (NF)-κB/p52 present with defects in humoral responses, germinal center reactions, and splenic microarchitecture. J. Exp. Med. 187:147–59 Zelazowski P, Carrasco D, Rosas FR, Moorman MA, Bravo R, Snapper CM. 1997. B cells genetically deficient in the c-Rel transactivation domain have selective defects in germline CH transcription and Ig class switching. J. Immunol. 159:3133–39 Carrasco D, Cheng J, Lewin A, Warr G, Yang H, Rizzo C, Rosas F, Snapper C, Bravo R. 1998. Multiple hemopoietic defects and lymphoid hyperplasia in mice lacking the transcriptional activation domain of the c-Rel protein. J. Exp. Med. 187:973–84 Franzoso G, Carlson L, Scharton-Kersten T, Shores EW, Epstein S, Grinberg A, Tran T, Shacter E, Leonardi A, Anver M, et al. 1997. Critical roles for the Bcl-3 oncoprotein in T cell-mediated immunity, splenic microarchitecture, and germinal center reactions. Immunity 6:479– 90 Toellner KM, Gulbranson Judge A, Taylor DR, Sze DM, MacLennan IC. 1996.

1 Feb 2002

10:12

184

Annu. Rev. Immunol. 2002.20:165-196. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

37a.

38.

39.

40.

41.

42.

43.

44.

AR

HONJO

AR152-07.tex

¥

AR152-07.SGM

KINOSHITA

¥

LaTeX2e(2001/05/10)

P1: GJC

MURAMATSU

Immunoglobulin switch transcript production in vivo related to the site and time of antigen-specific B cell activation. J. Exp. Med. 183:2303–12 Kinoshita K, Harigai H, Fagarasan S, Muramatsu M, Honjo T. 2001. A hallmark of active class switch recombination: transcripts directed by I promoters on looped-out circular DNAs. Proc. Natl. Acad. Sci. USA 98:12,620–23 Oliver AM, Martin F, Gartland GL, Carter RH, Kearney JF. 1997. Marginal zone B cells exhibit unique activation, proliferative and immunoglobulin secretory responses. Eur. J. Immunol. 27:2366– 74 Guinamard R, Okigaki M, Schlessinger J, Ravetch JV. 2000. Absence of marginal zone B cells in Pyk-2–deficient mice defines their role in the humoral response. Nat. Immunol. 1:31–36 Macpherson AJ, Gatto D, Sainsbury E, Harriman GR, Hengartner H, Zinkernagel RM. 2000. A primitive T cellindependent mechanism of intestinal mucosal IgA responses to commensal bacteria. Science 288:2222–26 Macpherson AJ, Lamarre A, McCoy K, Harriman GR, Odermatt B, Dougan G, Hengartner H, Zinkernagel RM. 2001. IgA production without µ or δ chain expression in developing B cells. Nat. Immunol. 2:625–31 de Waard R, Dammers PM, Tung JW, Kantor AB, Wilshire JA, Bos NA, Herzenberg LA, Kroese FG. 1998. Presence of germline and full-length IgA RNA transcripts among peritoneal B-1 cells. Dev. Immunol. 6:81–87 Kroese FG, Ammerlaan WA, Deenen GJ, Adams S, Herzenberg LA, Kantor AB. 1995. A dual origin for IgA plasma cells in the murine small intestine. Adv. Exp. Med. Biol. A 371:435–40 Kroese FG, Butcher EC, Stall AM, Herzenberg LA. 1989. A major peritoneal reservoir of precursors for intestinal IgA plasma cells. Immunol. Invest. 18:47–58

45. Bos NA, Bun JC, Popma SH, Cebra ER, Deenen GJ, van der Cammen MJ, Kroese FG, Cebra JJ. 1996. Monoclonal immunoglobulin A derived from peritoneal B cells is encoded by both germ line and somatically mutated VH genes and is reactive with commensal bacteria. Infect. Immun. 64:616–23 46. Fagarasan S, Kinoshita K, Muramatsu M, Ikuta K, Honjo T. 2001. In situ class switching and differentiation to IgA producing cells in the gut lamina propria. Nature. 413:639–43 47. Matsumoto M, Lo SF, Carruthers CJ, Min J, Mariathasan S, Huang G, Plas DR, Martin SM, Geha RS, Nahm MH, Chaplin DD. 1996. Affinity maturation without germinal centres in lymphotoxin-αdeficient mice. Nature 382:462–66 48. Kato J, Motoyama N, Taniuchi I, Takeshita H, Toyoda M, Masuda K, Watanabe T. 1998. Affinity maturation in Lyn kinase-deficient mice with defective germinal center formation. J. Immunol. 160: 4788–95 49. Weller S, Faili A, Garcia C, Braun MC, Le Deist FF, de Saint Basile GG, Hermine O, Fischer A, Reynaud C, Weill J. 2001. CD40–CD40L independent Ig gene hypermutation suggests a second B cell diversification pathway in humans. Proc. Natl. Acad. Sci. USA 98:1166– 70 50. Shimizu A, Takahashi N, Yaoita Y, Honjo T. 1982. Organization of the constantregion gene family of the mouse immunoglobulin heavy chain. Cell 28:499– 506 51. Kinoshita K, Honjo T. 2001. Linking class-switch recombination with somatic hypermutation. Nat. Rev. Mol. Cell Biol. 2:493–503 52. Iwasato T, Shimizu A, Honjo T, Yamagishi H. 1990. Circular DNA is excised by immunoglobulin class switch recombination. Cell 62:143–49 53. von Schwedler U, Jack HM, Wabl M. 1990. Circular DNA is a product of the

1 Feb 2002

10:12

AR

AR152-07.tex

AR152-07.SGM

LaTeX2e(2001/05/10)

P1: GJC

MECHANISM OF SWITCH RECOMBINATION

54.

Annu. Rev. Immunol. 2002.20:165-196. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

55.

56.

57.

58.

59.

60.

61.

62.

immunoglobulin class switch rearrangement. Nature 345:452–56 Matsuoka M, Yoshida K, Maeda T, Usuda S, Sakano H. 1990. Switch circular DNA formed in cytokine-treated mouse splenocytes: evidence for intramolecular DNA deletion in immunoglobulin class switching. Cell 62:135–42 Lorenz M, Jung S, Radbruch A. 1995. Switch transcripts in immunoglobulin class switching. Science 267:1825–28 Hein K, Lorenz MG, Siebenkotten G, Petry K, Christine R, Radbruch A. 1998. Processing of switch transcripts is required for targeting of antibody class switch recombination. J. Exp. Med. 188: 2369–74 Nikaido T, Yamawaki Kataoka Y, Honjo T. 1982. Nucleotide sequences of switch regions of immunoglobulin Cε and Cγ genes and their comparison. J. Biol. Chem. 257:7322–29 Nikaido T, Nakai S, Honjo T. 1981. Switch region of immunoglobulin Cµ gene is composed of simple tandem repetitive sequences. Nature 292:845–48 Arakawa H, Iwasato T, Hayashida H, Shimizu A, Honjo T, Yamagishi H. 1993. The complete murine immunoglobulin class switch region of the α heavy chain gene-hierarchic repetitive structure and recombination breakpoints. J. Biol. Chem. 268:4651–55 Takahashi N, Nakai S, Honjo T. 1980. Cloning of human immunoglobulin µ gene and comparison with mouse µ gene. Nucleic Acids Res. 8:5983–91 Mußmann R, Courtet M, Schwager J, Du Pasquier L. 1997. Microsites for immunoglobulin switch recombination breakpoints from Xenopus to mammals. Eur. J. Immunol. 27:2610–19 Kingzette M, Spieker-Polet H, Yam PC, Zhai SK, Knight KL. 1998. Trans-chromosomal recombination within the Ig heavy chain switch region in B lymphocytes. Proc. Natl. Acad. Sci. USA 95: 11840–45

185

63. Kitao H, Arakawa H, Yamagishi H, Shimizu A. 1996. Chicken immunoglobulin µ-chain gene: germline organization and tandem repeats characteristic of class switch recombination. Immunol. Lett. 52: 99–104 64. Kitao H, Arakawa H, Kuma K, Yamagishi H, Nakamura N, Furusawa S, Matsuda H, Yasuda M, Ekino S, Shimizu A. 2000. Class switch recombination of the chicken IgH chain genes: implications for the primordial switch region repeats. Int. Immunol. 12:959–68 65. Sun J, Butler JE. 1997. Sequence analysis of pig switch µ, Cµ, and Cµm. Immunogenetics 46:452–60 66. Nguyen VK, Hamers R, Wyns L, Muyldermans S. 1999. Loss of splice consensus signal is responsible for the removal of the entire C(H)1 domain of the functional camel IGG2A heavy-chain antibodies. Mol. Immunol. 36:515–24 67. Ishiguro H, Ichihara Y, Namikawa T, Nagatsu T, Kurosawa Y. 1989. Nucleotide sequence of Suncus murinus immunoglobulin µ gene and comparison with mouse and human µ genes. FEBS Lett. 247:317–22 68. Knight KL, Becker RS. 1987. Isolation of genes encoding bovine IgM, IgG, IgA and IgE chains. Vet. Immunol. Immunopathol. 17:17–24 69. Leung H, Maizels N. 1992. Transcriptional regulatory elements stimulate recombination in extrachromosomal substrates carrying immunoglobulin switchregion sequences. Proc. Natl. Acad. Sci. USA 89:4154–58 70. Lepse CL, Kumar R, Ganea D. 1994. Extrachromosomal eukaryotic DNA substrates for switch recombination: analysis of isotype and cell specificity. DNA Cell Biol. 13:1151–61 71. Daniels GA, Lieber MR. 1995. Strand specificity in the transcriptional targeting of recombination at immunoglobulin switch sequences. Proc. Natl. Acad. Sci. USA 92:5625–29

1 Feb 2002

10:12

Annu. Rev. Immunol. 2002.20:165-196. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

186

AR

HONJO

AR152-07.tex

¥

AR152-07.SGM

KINOSHITA

¥

LaTeX2e(2001/05/10)

P1: GJC

MURAMATSU

72. Ott DE, Marcu KB. 1989. Molecular requirements for immunoglobulin heavy chain constant region gene switch-recombination revealed with switch-substrate retroviruses. Int. Immunol. 1:582–91 73. Kinoshita K, Tashiro J, Tomita S, Lee CG, Honjo T. 1998. Target specificity of immunoglobulin class switch recombination is not determined by nucleotide sequences of S regions. Immunity 9:849– 58 74. Stavnezer J, Bradley SP, Rousseau N, Pearson T, Shanmugam A, Waite DJ, Rogers PR, Kenter AL. 1999. Switch recombination in a transfected plasmid occurs preferentially in a B cell line that undergoes switch recombination of its chromosomal Ig heavy chain genes. J. Immunol. 163:2028–40 75. Petry K, Siebenkotten G, Christine R, Hein K, Radbruch A. 1999. An extrachromosomal switch recombination substrate reveals kinetics and substrate requirements of switch recombination in primary murine B cells. Int. Immunol. 11:753–63 76. Takahashi N, Ueda S, Obata M, Nikaido T, Nakai S, Honjo T. 1982. Structure of human immunoglobulin γ genes: implications for evolution of a gene family. Cell 29:671–79 77. Kinoshita K, Shimizu A, Honjo T. 1991. The membrane exons of the pseudo-γ chain gene of the human immunoglobulin are apparently functional and highly homologous to those of the γ 1 gene. Immunol. Lett. 27:151–55 78. Luby TM, Schrader CE, Stavnezer J, Selsing E. 2001. The µ switch region tandem repeats are important, but not required, for antibody class switch recombination. J. Exp. Med. 193:159–68 79. Willers J, Kolb C, Weiler E. 1999. Apparent trans-chromosomal antibody class switch in mice bearing an Igh(a) µ-chain transgene on an Igh(b) genetic background. Immunobiology 200:150–64 80. Rolink A, Melchers F, Andersson J. 1996. The SCID but not the RAG-2 gene pro-

81.

82.

83.

84.

85.

86.

87.

88.

89.

duct is required for Sµ-Sε heavy chain class switching. Immunity 5:319–30 Manis JP, Gu Y, Lansford R, Sonoda E, Ferrini R, Davidson L, Rajewsky K, Alt FW. 1998. Ku70 is required for late B cell development and immunoglobulin heavy chain class switching. J. Exp. Med. 187:2081–89 Casellas R, Nussenzweig A, Wuerffel R, Pelanda R, Reichlin A, Suh H, Qin XF, Besmer E, Kenter A, Rajewsky K, Nussenzweig MC. 1998. Ku80 is required for immunoglobulin isotype switching. EMBO J. 17:2404–11 Dunnick W, Hertz GZ, Scappino L, Gritzmacher C. 1993. DNA sequences at immunoglobulin switch region recombination sites. Nucleic Acids Res. 21:365– 72 Lee C-G, Kondo S, Honjo T. 1998. Frequent but biased class switch recombination in the Sµ flanking regions. Curr. Biol. 8:227–30 Shanmugam A, Shi MJ, Yauch L, Stavnezer J, Kenter AL. 2000. Evidence for class-specific factors in immunoglobulin isotype switching. J. Exp. Med. 191: 1365–80 Pan Q, Hammarstrom L. 1999. Targeting of human switch recombination breakpoints: implications for the mechanism of µ-γ isotype switching. Eur. J. Immunol. 29:2779–87 Ott DE, Alt FW, Marcu KB. 1987. Immunoglobulin heavy chain switch region recombination within a retroviral vector in murine pre-B cells. EMBO J. 6:557– 84 Ballantyne J, Henry DL, Marcu KB. 1997. Antibody class switch recombinase activity is B cell stage specific and functions stochastically in the absence of ‘targeted accessibility’ control. Int. Immunol. 9:963–74 Gu H, Zou YR, Rajewsky K. 1993. Independent control of immunoglobulin switch recombination at individual switch regions evidenced through

1 Feb 2002

10:12

AR

AR152-07.tex

AR152-07.SGM

LaTeX2e(2001/05/10)

P1: GJC

MECHANISM OF SWITCH RECOMBINATION

90.

Annu. Rev. Immunol. 2002.20:165-196. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

91.

92.

93.

94.

95.

96.

97.

98.

Cre-loxP-mediated gene targeting. Cell 73:1155–64 Jung S, Rajewsky K, Radbruch A. 1993. Shutdown of class switch recombination by deletion of a switch region control element. Science 259:984–87 Zhang J, Bottaro A, Li S, Stewart V, Alt FW. 1993. A selective defect in IgG2b switching as a result of targeted mutation of the Iγ 2b promoter and exon. EMBO J. 12:3529–37 Seidl KJ, Bottaro A, Vo A, Zhang J, Davidson L, Alt FW. 1998. An expressed neo(r) cassette provides required functions of the Iγ 2b exon for class switching. Int. Immunol. 10:1683–92 Li MJ, Chung W, Maizels N. 1997. Developmental specificity of immunoglobulin heavy chain switch region recombination activities. Mol. Immunol. 34:201–8 Ballantyne J, Henry DL, Muller JR, Briere F, Snapper CM, Kehry M, Marcu KB. 1998. Efficient recombination of a switch substrate retrovector in CD40–activated B lymphocytes: implications for the control of CH gene switch recombination. J. Immunol. 161:1336–47 Stavnezer NJ, Sirlin S. 1986. Specificity of immunoglobulin heavy chain switch correlates with activity of germline heavy chain genes prior to switching. EMBO J. 5:95–102 Yancopoulos GD, DePinho RA, Zimmerman KA, Lutzker SG, Rosenberg N, Alt FW. 1986. Secondary genomic rearrangement events in pre-B cells: VHDJH replacement by a LINE-1 sequence and directed class switching. EMBO J. 5:3259– 66 Bottaro A, Lansford R, Xu L, Zhang J, Rothman P, Alt FW. 1994. S region transcription per se promotes basal IgE class switch recombination but additional factors regulate the efficiency of the process. EMBO J. 13:665–74 Bottaro A, Young F, Chen J, Serwe M, Sablitzky F, Alt FW. 1998. Deletion of the IgH intronic enhancer and associated

99.

100.

101.

102.

103.

104.

105.

106.

187

matrix-attachment regions decreases, but does not abolish, class switching at the µ locus. Int. Immunol. 10:799–806 Harriman GR, Bradley A, Das S, Rogers Fani P, Davis AC. 1996. IgA class switch in Iα exon-deficient mice. Role of germline transcription in class switch recombination. J. Clin. Invest. 97:477–85 Qiu G, Harriman GR, Stavnezer J. 1999. Iα exon-replacement mice synthesize a spliced HPRT-Cα transcript which may explain their ability to switch to IgA. Inhibition of switching to IgG in these mice. Int. Immunol. 11:37–46 Sakai E, Bottaro A, Alt FW. 1999. The Ig heavy chain intronic enhancer core region is necessary and sufficient to promote efficient class switch recombination. Int. Immunol. 11:1709–13 Seidl KJ, Manis JP, Bottaro A, Zhang J, Davidson L, Kisselgof A, Oettgen H, Alt FW. 1999. Position-dependent inhibition of class-switch recombination by PGK-neor cassettes inserted into the immunoglobulin heavy chain constant region locus. Proc. Natl. Acad. Sci. USA 96:3000–5 Xu L, Gorham B, Li SC, Bottaro A, Alt FW, Rothman P. 1993. Replacement of germ-line ε promoter by gene targeting alters control of immunoglobulin heavy chain class switching. Proc. Natl. Acad. Sci. USA 90:3705–9 Cogne M, Lansford R, Bottaro A, Zhang J, Gorman J, Young F, Cheng HL, Alt FW. 1994. A class switch control region at the 30 end of the immunoglobulin heavy chain locus. Cell 77:737–47 Manis JP, van der Stoep N, Tian M, Ferrini R, Davidson L, Bottaro A, Alt FW. 1998. Class switching in B cells lacking 30 immunoglobulin heavy chain enhancers. J. Exp. Med. 188:1421–31 Nakamura M, Kondo S, Sugai M, Nazarea M, Imamura S, Honjo T. 1996. High frequency class switching of an IgM+B lymphoma clone CH12F3 to IgA+ cells. Int. Immunol. 8:193–201

1 Feb 2002

10:12

Annu. Rev. Immunol. 2002.20:165-196. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

188

AR

HONJO

AR152-07.tex

¥

AR152-07.SGM

KINOSHITA

¥

LaTeX2e(2001/05/10)

P1: GJC

MURAMATSU

107. Muramatsu M, Sankaranand VS, Anant S, Sugai M, Kinoshita K, Davidson NO, Honjo T. 1999. Specific expression of activation-induced cytidine deaminase (AID), a novel member of the RNA-editing deaminase family in germinal center B cells. J. Biol. Chem. 274:18470–76 108. Lee C-G, Kinoshita K, Arudchandran A, Cerritelli SM, Crouch RJ, Honjo T. 2001. Quantitative regulation of class switch recombination by S region transcription. J. Exp. Med. 194:365–74 109. Zhang J, Alt FW, Honjo T. 1995. Regulation of class switch recombination of the immunoglobulin heavy chain genes. In Immunoglobulin Genes, ed. T Honjo, FW Alt, pp. 235–65. London: Academic 110. Ferlin WG, Severinson E, Strom L, Heath AW, Coffman RL, Ferrick DA, Howard MC. 1996. CD40 signaling induces interleukin-4–independent IgE switching in vivo. Eur J. Immunol. 26:2911–15 111. Ford GS, Yin CH, Barnhart B, Sztam K, Covey LR. 1998. CD40 ligand exerts differential effects on the expression of Iγ transcripts in subclones of an IgM+ human B cell lymphoma line. J. Immunol. 160:595–605 112. Fujieda S, Saxon A, Zhang K. 1996. Direct evidence that γ 1 and γ 3 switching in human B cells is interleukin-10 dependent. Mol. Immunol. 33:1335–43 113. Linehan LA, Warren WD, Thompson PA, Grusby MJ, Berton MT. 1998. STAT6 is required for IL-4–induced germline Ig gene transcription and switch recombination. J. Immunol. 161:302–10 114. Mizoguchi C, Uehara S, Akira S, Takatsu K. 1999. IL-5 induces IgG1 isotype switch recombination in mouse CD38–activated sIgD-positive B lymphocytes. J. Immunol. 162:2812–19 115. Tokuyama H, Tokuyama Y. 1997. Retinoic acid induces the expression of germline Cα transcript mainly by a TGF-βindependent mechanism. Cell. Immunol. 176:14–21 116. Yanagihara Y, Basaki Y, Ikizawa K, Kaji-

117.

118.

119.

120.

121.

122.

123.

wara K. 1997. Possible role of nuclear factor-κB activity in germline Cε transcription in a human Burkitt lymphoma B cell line. Cell. Immunol. 176:66–74 Zan H, Cerutti A, Dramitinos P, Schaffer A, Casali P. 1998. CD40 engagement triggers switching to IgA1 and IgA2 in human B cells through induction of endogenous TGF-β: evidence for TGF-β but not IL-10–dependent direct Sµ→Sα and sequential Sµ→Sγ , Sγ →Sα DNA recombination. J. Immunol. 161:5217– 25 Ezernieks J, Schnarr B, Metz K, Duschl A. 1996. The human IgE germline promoter is regulated by interleukin-4, interleukin-13, interferon-α and interferonγ via an interferon-γ -activated site and its flanking regions. Eur. J. Biochem. 240:667–73 Iciek LA, Delphin SA, Stavnezer J. 1997. CD40 cross-linking induces Ig ε germline transcripts in B cells via activation of NFκB: synergy with IL-4 induction. J. Immunol. 158:4769–79 Jumper MD, Fujita K, Lipsky PE, Meek K. 1996. A CD30 responsive element in the germline ε promoter that is distinct from and inhibitory to the CD40 response element. Mol. Immunol. 33:965–72 Lin SC, Stavnezer J. 1996. Activation of NF-κB/Rel by CD40 engagement induces the mouse germ line immunoglobulin Cγ 1 promoter. Mol. Cell. Biol. 16:4591– 603 Lin SC, Wortis HH, Stavnezer J. 1998. The ability of CD40L, but not lipopolysaccharide, to initiate immunoglobulin switching to immunoglobulin G1 is explained by differential induction of NFκB/Rel proteins. Mol. Cell. Biol. 18: 5523–32 Messner B, Stutz AM, Albrecht B, Peiritsch S, Woisetschlager M. 1997. Cooperation of binding sites for STAT6 and NFκB/rel in the IL-4–induced up-regulation of the human IgE germline promoter. J. Immunol. 159:3330–37

1 Feb 2002

10:12

AR

AR152-07.tex

AR152-07.SGM

LaTeX2e(2001/05/10)

P1: GJC

Annu. Rev. Immunol. 2002.20:165-196. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

MECHANISM OF SWITCH RECOMBINATION 124. Pan Q, Petit-Frere C, Stavnezer J, Hammarstrom L. 2000. Regulation of the promoter for human immunoglobulin γ 3 germ-line transcription and its interaction with the 30 α enhancer. Eur. J. Immunol. 30:1019–29 125. Pan Q, Petit-Frere C, Hammarstrom L. 2000. An allotype-associated polymorphism in the γ 3 promoter determines the germ-line γ 3 transcriptional rate but does not influence switching and subsequent IgG3 production. Eur. J. Immunol. 30:2388–93 126. Pardali E, Xie XQ, Tsapogas P, Itoh S, Arvanitidis K, Heldin CH, ten Dijke P, Grundstrom T, Sideras P. 2000. Smad and AML proteins synergistically confer transforming growth factor β1 responsiveness to human germ-line IgA genes. J. Biol. Chem. 275:3552–60 127. Strom L, Laurencikiene J, Miskiniene A, Severinson E. 1999. Characterization of CD40–dependent immunoglobulin class switching. Scand. J. Immunol. 49:523– 32 128. Warren WD, Roberts KL, Linehan LA, Berton MT. 1999. Regulation of the germline immunoglobulin Cγ 1 promoter by CD40 ligand and IL-4: dual role for tandem NF-κB binding sites. Mol. Immunol. 36:31–44 129. Xie XQ, Pardali E, Holm M, Sideras P, Grundstrom T. 1999. AML and Ets proteins regulate the Iα1 germ-line promoter. Eur. J. Immunol. 29:488–98 130. Sakai E, Bottaro A, Davidson L, Sleckman BP, Alt FW. 1999. Recombination and transcription of the endogenous Ig heavy chain locus is effected by the Ig heavy chain intronic enhancer core region in the absence of the matrix attachment regions. Proc. Natl. Acad. Sci. USA 96:1526–31 131. Radcliffe G, Lin Y-C, Julius M, Marcu KB, Stavnezer J. 1990. Structure of germ line immunoglobulin α heavy-chain RNA and its location on polysomes. Mol. Cell. Biol. 10:382–86

189

132. Bachl J, Turck CW, Wabl M. 1996. Translatable immunoglobulin germ-line transcript. Eur. J. Immunol. 26:870–74 133. Kinoshita K, Honjo T. 2000. Unique and unprecedented recombination mechanisms in class switching. Curr. Opin. Immunol. 12:195–98 134. Tashiro J, Kinoshita K, Honjo T. 2001. Palindromic but not G-rich sequences are targets of class switch recombination. Int. Immunol. 13:495–505 135. Storb U, Peters A, Klotz E, Kim N, Shen HM, Hackett J, Rogerson B, Martin TE. 1998. cis-acting sequences that affect somatic hypermutation of Ig genes. Immunol. Rev. 162:153–60 136. Gnatt AL, Cramer P, Fu J, Bushnell DA, Kornberg RD. 2001. Structural basis of transcription: an RNA polymerase ˚ resolution. II elongation complex at 3.3 A Science 292:1876–82 137. Kataoka T, Miyata T, Honjo T. 1981. Repetitive sequences in class-switch recombination regions of immunoglobulin heavy chain genes. Cell 23:357–68 138. Reaban ME, Griffin JA. 1990. Induction of RNA-stabilized DNA conformers by transcription of an immunoglobulin switch region. Nature 348:342–44 139. Daniels GA, Lieber MR. 1995. RNA: DNA complex formation upon transcription of immunoglobulin switch regions: implications for the mechanism and regulation of class switch recombination. Nucleic Acids Res. 23:5006–11 140. Tian M, Alt FW. 2000. Transcriptioninduced cleavage of immunoglobulin switch regions by nucleotide excision repair nucleases in vitro. J. Biol. Chem. 275:24163–72 141. Sen D, Gilbert W. 1988. Formation of parallel four-stranded complexes by guaninerich motifs in DNA and its implications for meiosis. Nature 334:364–66 142. McBlane JF, van Gent DC, Ramsden DA, Romeo C, Cuomo CA, Gellert M, Oettinger MA. 1995. Cleavage at a V(D)J recombination signal requires only RAG1

1 Feb 2002

10:12

190

143.

Annu. Rev. Immunol. 2002.20:165-196. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

144.

145.

146.

147.

148.

148a.

149.

AR

HONJO

AR152-07.tex

¥

AR152-07.SGM

KINOSHITA

¥

LaTeX2e(2001/05/10)

P1: GJC

MURAMATSU

and RAG2 proteins and occurs in two steps. Cell 83:387–95 Wuerffel RA, Du J, Thompson RJ, Kenter AL. 1997. Ig Sγ 3 DNA-specifc double strand breaks are induced in mitogenactivated B cells and are implicated in switch recombination. J. Immunol. 159:4139–44 Wuerffel R, Jamieson CE, Morgan L, Merkulov GV, Sen R, Kenter AL. 1992. Switch recombination breakpoints are strictly correlated with DNA recognition motifs for immunoglobulin Sγ 3 DNAbinding proteins. J. Exp. Med. 176:339– 49 Kenter AL, Wuerffel R, Sen R, Jamieson CE, Merkulov GV. 1993. Switch recombination breakpoints occur at nonrandom positions in the Sγ tandem repeat. J. Immunol. 151:4718–31 Du J, Zhu Y, Shanmugam A, Kenter AL. 1997. Analysis of immunoglobulin Sγ 3 recombination breakpoints by PCR: implications for the mechanism of isotype switching. Nucleic Acids Res. 25: 3066–73 Zhao Y, Rabbani H, Shimizu A, Hammarstrom L. 2000. Mapping of the chicken immunoglobulin heavy-chain constant region gene locus reveals an inverted α gene upstream of a condensed ε gene. Immunology 101:348–53 Kinoshita K, Lee C-G, Tashiro J, Muramatsu M, Chen X-C, Yoshikawa K, Honjo T. 1999. Molecular mechanism of immunoglobulin class switch recombination. Cold Spring Harbor Symp. Quantitative Biol. Signaling Gene Expression in the Immune System, pp. 217– 26: New York: Cold Spring Harbor Lab. Press Chen X, Kinoshita K, Honjo T. 2001. Variable deletion and duplication at recombination junction ends: implication for staggered double-strand cleavage in class switch recombination. Proc. Natl. Acad. Sci. USA In press Li MJ, Peakman MC, Golub EI, Reddy

150.

151.

152.

153.

154.

155.

156.

157.

G, Ward DC, Radding CM, Maizels N. 1996. Rad51 expression and localization in B cells carrying out class switch recombination. Proc. Natl. Acad. Sci. USA 93:10222–27 Li MJ, Maizels N. 1997. Nuclear Rad51 foci induced by DNA damage are distinct from Rad51 foci associated with B cell activation and recombination. Exp. Cell Res. 237:93–100 Peakman MC, Maizels N. 1998. Localization of splenic B cells activated for switch recombination by in situ hybridization with Iγ 1 switch transcript and Rad51 probes. J. Immunol. 161:4008–15 Essers J, Hendriks RW, Swagemakers SM, Troelstra C, de Wit J, Bootsma D, Hoeijmakers JH, Kanaar R. 1997. Disruption of mouse RAD54 reduces ionizing radiation resistance and homologous recombination. Cell 89:195–204 Kenter AL, Tredup J. 1991. High expression of a 30 —50 exonuclease activity is specific to B lymphocytes. Mol. Cell. Biol. 11:4398–404 Muller JR, Marcu KB. 1998. Stimulation of murine B lymphocytes induces a DNA exonuclease whose activity on switch-µ DNA is specifically inhibited by other germ-line switch region RNAs. J. Immunol. 160:3337–41 Dunnick W, Wilson M, Stavnezer J. 1989. Mutations, duplication, and deletion of recombined switch regions suggest a role for DNA replication in the immunoglobulin heavy-chain switch. Mol. Cell. Biol. 9:1850–56 Li J, Daniels GA, Lieber MR. 1996. Asymmetric mutation around the recombination break point of immunoglobulin class switch sequences on extrachromosomal substrates. Nucleic Acids Res. 24:2104–11 Friedberg EC, Feaver WJ, Gerlach VL. 2000. The many faces of DNA polymerases:strategies for mutagenesis and for mutational avoidance. Proc. Natl. Acad. Sci. USA 97:5681–83

1 Feb 2002

10:12

AR

AR152-07.tex

AR152-07.SGM

LaTeX2e(2001/05/10)

P1: GJC

Annu. Rev. Immunol. 2002.20:165-196. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

MECHANISM OF SWITCH RECOMBINATION 158. Kim N, Kage K, Matsuda F, Lefranc MP, Storb U. 1997. B lymphocytes of xeroderma pigmentosum or Cockayne syndrome patients with inherited defects in nucleotide excision repair are fully capable of somatic hypermutation of immunoglobulin genes. J. Exp. Med. 186:413–19 159. Shen HM, Cheo DL, Friedberg E, Storb U. 1997. The inactivation of the XP-C gene does not affect somatic hypermutation or class switch recombination of immunoglobulin genes. Mol. Immunol. 34:527–33 160. Ehrenstein MR, Neuberger MS. 1999. Deficiency in Msh2 affects the efficiency and local sequence specificity of immunoglobulin class-switch recombination: parallels with somatic hypermutation. EMBO J. 18:3484–90 161. Schrader CE, Edelmann W, Kucherlapati R, Stavnezer J. 1999. Reduced isotype switching in splenic B cells from mice deficient in mismatch repair enzymes. J. Exp. Med. 190:323–30 162. Mizuta TR, Fukita Y, Miyoshi T, Shimizu A, Honjo T. 1993. Isolation of cDNA encoding a binding protein specific to 50 -phosphorylated singlestranded DNA with G-rich sequences. Nucleic Acids Res. 21:1761–66 163. Zhang Q, Wang YC, Montalvo EA. 1999. Sµbp-2 represses the Epstein-Barr virus lytic switch promoter. Virology 255:160–70 164. Cox GA, Mahaffey CL, Frankel WN. 1998. Identification of the mouse neuromuscular degeneration gene and mapping of a second site suppressor allele. Neuron 21:1327–37 165. Miao M, Chan SL, Fletcher GL, Hew CL. 2000. The rat ortholog of the presumptive flounder antifreeze enhancerbinding protein is a helicase domaincontaining protein. Eur. J. Biochem. 267:7237–46 166. Sebastiani G, Durocher D, Gros P, Nemer M, Malo D. 1995. Localization of

167.

168.

169.

170.

171.

172.

173.

174.

175.

191

the Catf1 transcription factor gene to mouse chromosome 19. Mamm. Genome 6:147–48 Schultz CL, Elenich LA, Dunnick WA. 1991. Nuclear protein binding to octamer motifs in the immunoglobulin γ 1 switch region. Int. Immunol. 3:109–16 Waters SH, Saikh KU, Stavnezer J. 1989. A B-cell-specific nuclear protein that binds to DNA sites 50 to immunoglobulin Sα tandem repeats is regulated during differentiation. Mol. Cell. Biol. 9:5594–601 Ma L, Hu B, Kenter AL. 1997. Ig Sγ specific DNA binding protein SNAP is related to the helix-loop-helix transcription factor E47. Int. Immunol. 9:1021–29 Xu L, Kim MG, Marcu KB. 1992. Properties of B cell stage specific and ubiquitous nuclear factors binding to immunoglobulin heavy chain gene switch regions. Int. Immunol. 4:875–87 Miranda GA, Chokler I, Aguilera RJ. 1995. The murine nucleolin protein is an inducible DNA and ATP binding protein which is readily detected in nuclear extracts of lipopolysaccharidetreated splenocytes. Exp. Cell Res. 217: 294–308 Goldfarb AN, Flores JP, Lewandowska K. 1996. Involvement of the E2A basic helix-loop-helix protein in immunoglobulin heavy chain class switching. Mol. Immunol. 33:947–56 Quong MW, Harris DP, Swain SL, Murre C. 1999. E2A activity is induced during B-cell activation to promote immunoglobulin class switch recombination. EMBO J. 18:6307–18 Williams M, Maizels N. 1991. LR1, a lipopolysaccharide-responsive factor with binding sites in the immunoglobulin switch regions and heavy-chain enhancer. Genes Dev. 5:2353–61 Hanakahi LA, Dempsey LA, Li MJ, Maizels N. 1997. Nucleolin is one component of the B cell-specific transcription factor and switch region binding

1 Feb 2002

10:12

192

176.

Annu. Rev. Immunol. 2002.20:165-196. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

177.

178.

179.

180.

181.

182.

183.

184.

AR

HONJO

AR152-07.tex

¥

AR152-07.SGM

KINOSHITA

¥

LaTeX2e(2001/05/10)

P1: GJC

MURAMATSU

protein, LR1. Proc. Natl. Acad. Sci. USA 94:3605–10 Dempsey LA, Hanakahi LA, Maizels N. 1998. A specific isoform of hnRNP D interacts with DNA in the LR1 heterodimer: canonical RNA binding motifs in a sequence-specific duplex DNA binding protein. J. Biol. Chem. 273:29224–29 Hanakahi LA, Maizels N. 2000. Transcriptional activation by LR1 at the Eµ enhancer and switch region sites. Nucleic Acids Res. 28:2651–57 Chung IK, Mehta VB, Spitzner JR, Muller MT. 1992. Eukaryotic topoisomerase II cleavage of parallel stranded DNA tetraplexes. Nucleic Acids Res. 20: 1973–77 Borggrefe T, Wabl M, Akhmedov AT, Jessberger R. 1998. A B-cell-specific DNA recombination complex. J. Biol. Chem. 273:17025–35 Borggrefe T, Masat L, Wabl M, Riwar B, Cattoretti G, Jessberger R. 1999. Cellular, intracellular, and developmental expression patterns of murine SWAP70. Eur. J. Immunol. 29:1812–22 Masat L, Caldwell J, Armstrong R, Khoshnevisan H, Jessberger R, Herndier B, Wabl M, Ferrick D. 2000. Association of SWAP-70 with the B cell antigen receptor complex. Proc. Natl. Acad. Sci. USA 97:2180–84 Lyon CJ, Miranda GA, Piao JS, Aguilera RJ. 1996. Characterization of an endonuclease activity which preferentially cleaves the G-rich immunoglobulin switch repeat sequences. Mol. Immunol. 33:157–69 Lyon CJ, Aguilera RJ. 1997. Purification and characterization of the immunoglobulin switch sequence-specific endonuclease (Endo-SR) from bovine spleen. Mol. Immunol. 34:209–19 Wu TT, Kabat EA, Bilofsky H. 1979. Some sequence similarities among cloned mouse DNA segments that code for λ and κ light chains of immunoglobulins. Proc. Natl. Acad. Sci. USA 76:4617–21

185. Bachl J, Steinberg C, Wabl M. 1997. Critical test of hot spot motifs for immunoglobulin hypermutation. Eur. J. Immunol. 27:3398–403 186. Dorner T, Brezinschek HP, Brezinschek RI, Foster SJ, Domiati-Saad R, Lipsky PE. 1997. Analysis of the frequency and pattern of somatic mutations within nonproductively rearranged human variable heavy chain genes. J. Immunol. 158:2779–89 187. Rogozin IB, Kolchanov NA. 1992. Somatic hypermutagenesis in immunoglobulin genes. II. Influence of neighbouring base sequences on mutagenesis. Biochim. Biophys. Acta 1171:11–18 188. Foster SJ, Dorner T, Lipsky PE. 1999. Somatic hypermutation of VκJκ rearrangements: targeting of RGYW motifs on both DNA strands and preferential selection of mutated codons within RGYW motifs. Eur. J. Immunol. 29:4011–21 189. Monson NL, Dorner T, Lipsky PE. 2000. Targeting and selection of mutations in human Vλ rearrangements. Eur. J. Immunol. 30:1597–605 190. Lebecque SG, Gearhart PJ. 1990. Boundaries of somatic mutation in rearranged immunoglobulin genes: 50 boundary is near the promoter, and 30 boundary is approximately 1 kb from V(D)J gene. J. Exp. Med. 172:1717–27 191. Betz AG, Rada C, Pannell R, Milstein C, Neuberger MS. 1993. Passenger transgenes reveal intrinsic specificity of the antibody hypermutation mechanism: clustering, polarity, and specific hot spots. Proc. Natl. Acad. Sci. USA 90: 2385–88 192. Dorner T, Foster SJ, Farner NL, Lipsky PE. 1998. Somatic hypermutation of human immunoglobulin heavy chain genes: targeting of RGYW motifs on both DNA strands. Eur. J. Immunol. 28:3384–96 193. Goyenechea B, Milstein C. 1996. Modifying the sequence of an immunoglobulin V-gene alters the resulting pattern

1 Feb 2002

10:12

AR

AR152-07.tex

AR152-07.SGM

LaTeX2e(2001/05/10)

P1: GJC

MECHANISM OF SWITCH RECOMBINATION

194.

Annu. Rev. Immunol. 2002.20:165-196. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

195.

196.

197.

198.

199.

200.

201.

202.

of hypermutation. Proc. Natl. Acad. Sci. USA 93:13979–84 Oprea M, Cowell LG, Kepler TB. 2001. The targeting of somatic hypermutation closely resembles that of meiotic mutation. J. Immunol. 166:892–99 Pasqualucci L, Migliazza A, Fracchiolla N, William C, Neri A, Baldini L, Chaganti RS, Klein U, Kuppers R, Rajewsky K, Dalla-Favera R. 1998. BCL-6 mutations in normal germinal center B cells: evidence of somatic hypermutation acting outside Ig loci. Proc. Natl. Acad. Sci. USA 95:11816–21 Shen HM, Peters A, Baron B, Zhu X, Storb U. 1998. Mutation of BCL-6 gene in normal B cells by the process of somatic hypermutation of Ig genes. Science 280:1750–52 Muschen M, Re D, Jungnickel B, Diehl V, Rajewsky K, Kuppers R. 2000. Somatic mutation of the CD95 gene in human B cells as a side-effect of the germinal center reaction. J. Exp. Med. 192:1833–40 Pasqualucci L, Neumeister P, Goossens T, Nanjangud G, Chaganti RS, Kuppers R, Dalla-Favera R. 2001. Hypermutation of multiple proto-oncogenes in Bcell diffuse large-cell lymphomas. Nature 412:341–46 Rabbitts TH, Forster A, Baer R, Hamlyn PH. 1983. Transcription enhancer identified near the human Cµ immunoglobulin heavy chain gene is unavailable to the translocated c-myc gene in a Burkitt lymphoma. Nature 306:806–9 Taub R, Kelly K, Battey J, Latt S, Lenoir GM, Tantravahi U, Tu Z, Leder P. 1984. A novel alteration in the structure of an activated c-myc gene in a variant t(2;8) Burkitt lymphoma. Cell 37:511–20 Klotz EL, Hackett J, Storb U. 1998. Somatic hypermutation of an artificial test substrate within an Igκ transgene. J. Immunol. 161:782–90 Peters A, Storb U. 1996. Somatic hy-

203.

204.

205.

206.

207. 208.

209.

210.

211.

193

permutation of immunoglobulin genes is linked to transcription initiation. Immunity 4:57–65 Yelamos J, Klix N, Goyenechea B, Lozano F, Chui YL, Gonzalez Fernandez A, Pannell R, Neuberger MS, Milstein C. 1995. Targeting of non-Ig sequences in place of the V segment by somatic hypermutation. Nature 376:225– 29 Azuma T, Motoyama N, Fields LE, Loh DY. 1993. Mutations of the chloramphenicol acetyl transferase transgene driven by the immunoglobulin promoter and intron enhancer. Int. Immunol. 5:121–30 Storb U, Klotz EL, Hackett J, Kage K, Bozek G, Martin TE. 1998. A hypermutable insert in an immunoglobulin transgene contains hotspots of somatic mutation and sequences predicting highly stable structures in the RNA transcript. J. Exp. Med. 188:689–98 Kolchanov NA, Solovyov VV, Rogozin IB. 1987. Peculiarities of immunoglobulin gene structures as a basis for somatic mutation emergence. FEBS Lett. 214:87–91 Brenner S, Milstein C. 1966. Origin of antibody variation. Nature 211:242–43 Tumas-Brundage K, Manser T. 1997. The transcriptional promoter regulates hypermutation of the antibody heavy chain locus. J. Exp. Med. 185:239–50 Fukita Y, Jacobs H, Rajewsky K. 1998. Somatic hypermutation in the heavy chain locus correlates with transcription. Immunity 9:105–14 Bachl J, Carlson C, Gray-Schopfer V, Dessing M, Olsson C. 2001. Increased transcription levels induce higher mutation rates in a hypermutating cell line. J. Immunol. 166:5051–57 Jolly C, Neuberger M. 2001. Somatic hypermutation of immunoglobulin κ transgenes: association of mutability with demethylation. Immunol. Cell Biol. 79:18–22

1 Feb 2002

10:12

Annu. Rev. Immunol. 2002.20:165-196. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

194

AR

HONJO

AR152-07.tex

¥

AR152-07.SGM

KINOSHITA

¥

LaTeX2e(2001/05/10)

P1: GJC

MURAMATSU

212. Lichtenstein M, Keini G, Cedar H, Bergman Y. 1994. B cell-specific demethylation: a novel role for the intronic κ chain enhancer sequence. Cell 76:913– 23 213. Rada C, Yelamos J, Dean W, Milstein C. 1997. The 50 hypermutation boundary of κ chains is independent of local and neighbouring sequences and related to the distance from the initiation of transcription. Eur. J. Immunol. 27:3115–20 214. Winter DB, Sattar N, Mai JJ, Gearhart PJ. 1997. Insertion of 2 kb of bacteriophage DNA between an immunoglobulin promoter and leader exon stops somatic hypermutation in a κ transgene. Mol. Immunol. 34:359–66 215. Bachl J, Olsson C, Chitkara N, Wabl M. 1998. The Ig mutator is dependent on the presence, position, and orientation of the large intron enhancer. Proc. Natl. Acad. Sci. USA 95:2396–99 216. Betz AG, Milstein C, Gonzalez-Fernandez A, Pannell R, Larson T, Neuberger MS. 1994. Elements regulating somatic hypermutation of an immunoglobulin κ gene: critical role for the intron enhancer/matrix attachment region. Cell 77:239–48 217. Goyenechea B, Klix N, Yelamos J, Williams GT, Riddell A, Neuberger MS, Milstein C. 1997. Cells strongly expressing Igκ transgenes show clonal recruitment of hypermutation: a role for both MAR and the enhancers. EMBO J. 16:3987–94 218. Klix N, Jolly CJ, Davies SL, Bruggemann M, Williams GT, Neuberger MS. 1998. Multiple sequences from downstream of the Jκ cluster can combine to recruit somatic hypermutation to a heterologous, upstream mutation domain. Eur. J. Immunol. 28:317–26 219. van der Stoep N, Gorman JR, Alt FW. 1998. Reevaluation of 30 Eκ function in stage- and lineage-specific rearrangement and somatic hypermutation. Immunity 8:743–50

220. Sale JE, Neuberger MS. 1998. TdTaccessible breaks are scattered over the immunoglobulin V domain in a constitutively hypermutating B cell line. Immunity 9:859–69 221. Goossens T, Klein U, Kuppers R. 1998. Frequent occurrence of deletions and duplications during somatic hypermutation: implications for oncogene translocations and heavy chain disease. Proc. Natl. Acad. Sci. USA 95:2463–68 222. Wilson PC, de Bouteiller O, Liu YJ, Potter K, Banchereau J, Capra JD, Pascual V. 1998. Somatic hypermutation introduces insertions and deletions into immunoglobulin V genes. J. Exp. Med. 187:59–70 223. Bross L, Fukita Y, McBlane F, Demolliere C, Rajewsky K, Jacobs H. 2000. DNA double-strand breaks in immunoglobulin genes undergoing somatic hypermutation. Immunity 13:589–97 224. Papavasiliou FN, Schatz DG. 2000. Cell-cycle-regulated DNA doublestranded breaks in somatic hypermutation of immunoglobulin genes. Nature 408:216–21 225. Sonoda E, Sasaki MS, Buerstedde JM, Bezzubova O, Shinohara A, Ogawa H, Takata M, Yamaguchi-Iwai Y, Takeda S. 1998. Rad51–deficient vertebrate cells accumulate chromosomal breaks prior to cell death. EMBO J. 17:598–608 226. Kong Q, Maizels N. 2001. DNA breaks in hypermutating immunoglobulin genes. Evidence for a break-and-repair pathway of somatic hypermutation. Genetics 158:369–78 227. Bertocci B, Quint L, Delbos F, Garcia C, Reynaud CA, Weill JC. 1998. Probing immunoglobulin gene hypermutation with microsatellites suggests a nonreplicative short patch DNA synthesis process. Immunity 9:257–65 228. Wiesendanger M, Kneitz B, Edelmann W, Scharff MD. 2000. Somatic hypermutation in MutS homologue MSH3–, MSH6–, and MSH3/MSH6–deficient

1 Feb 2002

10:12

AR

AR152-07.tex

AR152-07.SGM

LaTeX2e(2001/05/10)

P1: GJC

MECHANISM OF SWITCH RECOMBINATION

229.

Annu. Rev. Immunol. 2002.20:165-196. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

230.

231.

232.

233.

234.

235.

236.

mice reveals a role for the MSH2–MSH6 heterodimer in modulating the base substitution pattern. J. Exp. Med. 191:579– 84 Kong Q, Maizels N. 1999. PMS2–deficiency diminishes hypermutation of a λ1 transgene in young but not older mice. Mol. Immunol. 36:83–91 Phung QH, Winter DB, Alrefai R, Gearhart PJ. 1999. Hypermutation in IgV genes from mice deficient in the MLH1 mismatch repair protein. J. Immunol. 162:3121–24 Winter DB, Phung QH, Umar A, Baker SM, Tarone RE, Tanaka K, Liskay RM, Kunkel TA, Bohr VA, Gearhart PJ. 1998. Altered spectra of hypermutation in antibodies from mice deficient for the DNA mismatch repair protein PMS2. Proc. Natl. Acad. Sci. USA 95:6953–58 Jacobs H, Fukita Y, van der Horst GT, de Boer J, Weeda G, Essers J, de Wind N, Engelward BP, Samson L, Verbeek S, et al. 1998. Hypermutation of immunoglobulin genes in memory B cells of DNA repair-deficient mice. J. Exp. Med. 187:1735–43 Rada C, Ehrenstein MR, Neuberger MS, Milstein C. 1998. Hot spot focusing of somatic hypermutation in MSH2–deficient mice suggests two stages of mutational targeting. Immunity 9:135–41 Frey S, Bertocci B, Delbos F, Quint L, Weill JC, Reynaud CA. 1998. Mismatch repair deficiency interferes with the accumulation of mutations in chronically stimulated B cells and not with the hypermutation process. Immunity 9:127–34 Phung QH, Winter DB, Cranston A, Tarone RE, Bohr VA, Fishel R, Gearhart PJ. 1998. Increased hypermutation at G and C nucleotides in immunoglobulin variable genes from mice deficient in the MSH2 mismatch repair protein. J. Exp. Med. 187:1745–51 Cascalho M, Wong J, Steinberg C, Wabl M. 1998. Mismatch repair co-opted by hypermutation. Science 279:1207–10

195

237. Vora KA, Tumas-Brundage KM, Lentz VM, Cranston A, Fishel R, Manser T. 1999. Severe attenuation of the B cell immune response in Msh2–deficient mice. J. Exp. Med. 189:471–82 238. Kim N, Bozek G, Lo JC, Storb U. 1999. Different mismatch repair deficiencies all have the same effects on somatic hypermutation: intact primary mechanism accompanied by secondary modifications. J. Exp. Med. 190:21– 30 239. Bemark M, Sale JE, Kim HJ, Berek C, Cosgrove RA, Neuberger MS. 2000. Somatic hypermutation in the absence of DNA-dependent protein kinase catalytic subunit (DNA-PKcs) or recombinationactivating gene (RAG)1 activity. J. Exp. Med. 192:1509–14 240. Zan H, Komori A, Li Z, Cerutti A, Schaffer A, Flajnik MF, Diaz M, Casali P. 2001. The translesion DNA polymerase ζ plays a major role in Ig and bcl-6 somatic hypermutation. Immunity 14:643–53 241. Zeng X, Winter DB, Kasmer C, Kraemer KH, Lehmann AR, Gearhart PJ. 2001. DNA polymerase η is an A-T mutator in somatic hypermutation of immunoglobulin variable genes. Nat. Immunol. 2:537–41 242. Rogozin IB, Pavlov YI, Bebenek K, Matsuda T, Kunkel TA. 2001. Somatic mutation hotspots correlate with DNA polymerase η error spectrum. Nat. Immunol. 2:530–36 243. Diaz M, Verkoczy LK, Flajnik MF, Klinman NR. 2001. Decreased frequency of somatic hypermutation and impaired affinity maturation but intact germinal center formation in mice expressing antisense RNA to DNA polymerase ζ . J. Immunol. 167:327–35 244. Johnson RE, Washington MT, Prakash S, Prakash L. 2000. Fidelity of human DNA polymerase η. J. Biol. Chem. 275: 7447–50 245. Matsuda T, Bebenek K, Masutani C,

1 Feb 2002

10:12

196

246.

Annu. Rev. Immunol. 2002.20:165-196. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

247.

248.

249.

250.

AR

HONJO

AR152-07.tex

¥

AR152-07.SGM

KINOSHITA

¥

LaTeX2e(2001/05/10)

P1: GJC

MURAMATSU

Hanaoka F, Kunkel TA. 2000. Low fidelity DNA synthesis by human DNA polymerase-η. Nature 404:1011–13 Masutani C, Kusumoto R, Iwai S, Hanaoka F. 2000. Mechanisms of accurate translesion synthesis by human DNA polymerase η. EMBO J. 19:3100– 9 Callard RE, Armitage RJ, Fanslow WC, Spriggs MK. 1993. CD40 ligand and its role in X-linked hyper-IgM syndrome. Immunol. Today 14:559–64 Muto T, Muramatsu M, Taniwaki M, Kinoshita K, Honjo T. 2000. Isolation, tissue distribution and chromosomal localization of the human activationinduced cytidine deaminase (hAID) gene. Genomics 68:85–88 Minegishi Y, Lavoie A, CunninghamRundles C, Bedard PM, Hebert J, Cote L, Dan K, Sedlak D, Buckley RH, Fischer A, et al. 2000. Mutations in activation-induced cytidine deaminase in patients with hyper IgM syndrome. Clin. Immunol. 97:203–10 Mehta A, Kinter MT, Sherman NE, Driscoll DM. 2000. Molecular cloning of apobec-1 complementation factor, a

251.

252.

253.

254.

novel RNA-binding protein involved in the editing of apolipoprotein B mRNA. Mol. Cell Biol. 20:1846–54 Lellek H, Kirsten R, Diehl I, Apostel F, Buck F, Greeve J. 2000. Purification and molecular cloning of a novel essential component of the apolipoprotein B mRNA editing enzyme-complex. J. Biol. Chem. 275:19848–56 Jacobs H, Bross L. 2001. Towards an understanding of somatic hypermutation. Curr. Opin. Immunol. 13:208–18 Takata M, Sasaki MS, Sonoda E, Morrison C, Hashimoto M, Utsumi H, Yamaguchi-Iwai Y, Shinohara A, Takeda S. 1998. Homologous recombination and non-homologous end-joining pathways of DNA double-strand break repair have overlapping roles in the maintenance of chromosomal integrity in vertebrate cells. EMBO J. 17:5497–508 Lundgren M, Strom L, Bergquist LO, Skog S, Heiden T, Stavnezer J, Severinson E. 1995. Cell cycle regulation of immunoglobulin class switch recombination and germ-line transcription: potential role of Ets family members. Eur. J. Immunol. 25:2042–51

Annu. Rev. Immunol. 2002.20:165-196. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

6 Feb 2002

8:18

AR

AR152-07-COLOR.tex

AR152-07-COLOR.SGM

LaTeX2e(2002/01/18)

P1: GDL

Figure 1 Model for the regulation of class-switch recombination and somatic hypermutation by AID. V and S regions are indicated by a red rectangle, and a green or blue oval, respectively, superimposed with their predicted secondary structures indicated by half circles. AID probably edits pre-mRNA for an unknown enzyme (dark blue scissors), converting it to mRNA for a nicking endonuclease specific to the stem-loop structure (open-mouth scissors). Nicking of S-region DNA occurs frequently at different positions on both strands of DNA, giving rise to double-stranded cleavages with staggered ends. Nicking of V-region DNA is less frequent, generating mostly singlestranded breaks. Single strand tails are either digested by exonuclease(s) (yellow pacman) or double-stranded by error-prone DNA polymerases (yellow circle), giving rise to deletion (red triangles), duplication ( pink bar) and/or mutations (white triangles) near cleavage sites. Processed ends in S regions are ligated by the non-homologous end-joining (NHEJ) pathway, whereas those in V regions are repaired by an unknown mechanism other than NHEJ.

P1: FRK

February 1, 2002

12:47

Annual Reviews

AR152-FM

Annual Review of Immunology Volume 20, 2002

Annu. Rev. Immunol. 2002.20:165-196. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

CONTENTS FRONTISPIECE, Charles A. Janeway, Jr. A TRIP THROUGH MY LIFE WITH AN IMMUNOLOGICAL THEME, Charles A. Janeway, Jr.

xiv 1

THE B7 FAMILY OF LIGANDS AND ITS RECEPTORS: NEW PATHWAYS FOR COSTIMULATION AND INHIBITION OF IMMUNE RESPONSES, Beatriz M. Carreno and Mary Collins

29

MAP KINASES IN THE IMMUNE RESPONSE, Chen Dong, Roger J. Davis, and Richard A. Flavell

PROSPECTS FOR VACCINE PROTECTION AGAINST HIV-1 INFECTION AND AIDS, Norman L. Letvin, Dan H. Barouch, and David C. Montefiori T CELL RESPONSE IN EXPERIMENTAL AUTOIMMUNE ENCEPHALOMYELITIS (EAE): ROLE OF SELF AND CROSS-REACTIVE ANTIGENS IN SHAPING, TUNING, AND REGULATING THE AUTOPATHOGENIC T CELL REPERTOIRE, Vijay K. Kuchroo, Ana C. Anderson, Hanspeter Waldner, Markus Munder, Estelle Bettelli, and Lindsay B. Nicholson

55 73

101

NEUROENDOCRINE REGULATION OF IMMUNITY, Jeanette I. Webster, Leonardo Tonelli, and Esther M. Sternberg

125

MOLECULAR MECHANISM OF CLASS SWITCH RECOMBINATION: LINKAGE WITH SOMATIC HYPERMUTATION, Tasuku Honjo, Kazuo Kinoshita, and Masamichi Muramatsu

165

INNATE IMMUNE RECOGNITION, Charles A. Janeway, Jr. and Ruslan Medzhitov

KIR: DIVERSE, RAPIDLY EVOLVING RECEPTORS OF INNATE AND ADAPTIVE IMMUNITY, Carlos Vilches and Peter Parham ORIGINS AND FUNCTIONS OF B-1 CELLS WITH NOTES ON THE ROLE OF CD5, Robert Berland and Henry H. Wortis E PROTEIN FUNCTION IN LYMPHOCYTE DEVELOPMENT, Melanie W. Quong, William J. Romanow, and Cornelis Murre

197 217 253 301

LYMPHOCYTE-MEDIATED CYTOTOXICITY, John H. Russell and Timothy J. Ley

SIGNAL TRANSDUCTION MEDIATED BY THE T CELL ANTIGEN x

323

P1: FRK

Annu. Rev. Immunol. 2002.20:165-196. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

February 1, 2002

12:47

Annual Reviews

AR152-FM

CONTENTS

xi

RECEPTOR: THE ROLE OF ADAPTER PROTEINS, Lawrence E. Samelson INTERACTION OF HEAT SHOCK PROTEINS WITH PEPTIDES AND ANTIGEN PRESENTING CELLS: CHAPERONING OF THE INNATE AND ADAPTIVE IMMUNE RESPONSES, Pramod Srivastava CHROMATIN STRUCTURE AND GENE REGULATION IN THE IMMUNE SYSTEM, Stephen T. Smale and Amanda G. Fisher PRODUCING NATURE’S GENE-CHIPS: THE GENERATION OF PEPTIDES FOR DISPLAY BY MHC CLASS I MOLECULES, Nilabh Shastri, Susan

371

Schwab, and Thomas Serwold

395 427

463

THE IMMUNOLOGY OF MUCOSAL MODELS OF INFLAMMATION, Warren Strober, Ivan J. Fuss, and Richard S. Blumberg T CELL MEMORY, Jonathan Sprent and Charles D. Surh

GENETIC DISSECTION OF IMMUNITY TO MYCOBACTERIA: THE HUMAN MODEL, Jean-Laurent Casanova and Laurent Abel ANTIGEN PRESENTATION AND T CELL STIMULATION BY DENDRITIC CELLS, Pierre Guermonprez, Jenny Valladeau, Laurence Zitvogel, Clotilde Th´ery, and Sebastian Amigorena

495 551 581

621

NEGATIVE REGULATION OF IMMUNORECEPTOR SIGNALING, Andr´e Veillette, Sylvain Latour, and Dominique Davidson

669

CPG MOTIFS IN BACTERIAL DNA AND THEIR IMMUNE EFFECTS, Arthur M. Krieg

PROTEIN KINASE Cθ

709 IN

T CELL ACTIVATION, Noah Isakov and Amnon

Altman

RANK-L AND RANK: T CELLS, BONE LOSS, AND MAMMALIAN EVOLUTION, Lars E. Theill, William J. Boyle, and Josef M. Penninger PHAGOCYTOSIS OF MICROBES: COMPLEXITY IN ACTION, David M. Underhill and Adrian Ozinsky

761 795 825

STRUCTURE AND FUNCTION OF NATURAL KILLER CELL RECEPTORS: MULTIPLE MOLECULAR SOLUTIONS TO SELF, NONSELF DISCRIMINATION, Kannan Natarajan, Nazzareno Dimasi, Jian Wang, Roy A. Mariuzza, and David H. Margulies

853

INDEXES Subject Index Cumulative Index of Contributing Authors, Volumes 1–20 Cumulative Index of Chapter Titles, Volumes 1–20

ERRATA An online log of corrections to Annual Review of Immunology chapters (if any, 1997 to the present) may be found at http://immunol.annualreviews.org/

887 915 925

8 Feb 2002

9:41

AR

AR152-08.tex

AR152-08.SGM

LaTeX2e(2001/05/10) P1: GJC 10.1146/annurev.immunol.20.083001.084359

Annu. Rev. Immunol. 2002. 20:197–216 DOI: 10.1146/annurev.immunol.20.083001.084359 c 2002 by Annual Reviews. All rights reserved Copyright °

INNATE IMMUNE RECOGNITION Charles A. Janeway, Jr. and Ruslan Medzhitov Annu. Rev. Immunol. 2002.20:197-216. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

Section of Immunobiology and Howard Hughes Medical Institute, Yale University School of Medicine, New Haven, Connecticut 06520-8011; e-mail: [email protected]; [email protected]

Key Words Toll, Drosophila, pathogen, pattern recognition, receptors ■ Abstract The innate immune system is a universal and ancient form of host defense against infection. Innate immune recognition relies on a limited number of germline-encoded receptors. These receptors evolved to recognize conserved products of microbial metabolism produced by microbial pathogens, but not by the host. Recognition of these molecular structures allows the immune system to distinguish infectious nonself from noninfectious self. Toll-like receptors play a major role in pathogen recognition and initiation of inflammatory and immune responses. Stimulation of Toll-like receptors by microbial products leads to the activation of signaling pathways that result in the induction of antimicrobial genes and inflammatory cytokines. In addition, stimulation of Toll-like receptors triggers dendritic cell maturation and results in the induction of costimulatory molecules and increased antigen-presenting capacity. Thus, microbial recognition by Toll-like receptors helps to direct adaptive immune responses to antigens derived from microbial pathogens.

INTRODUCTION Innate immunity covers many areas of host defense against pathogenic microbes, including the recognition of pathogen-associated molecular patterns (PAMPs) (1). In vertebrates, which are the only phylum that can mount an adaptive immune response, there are also mechanisms to inhibit the activation of innate immunity. An example is the inhibition of killing by natural killer (NK) cells, which are known to receive an inhibitory stimulus from MHC class I molecules. We concentrate in this review on the mechanisms of recognition that are truly innate, such that the genes are encoded in the germline DNA and do not require the gene rearrangement essential to adaptive immune recognition (Table 1). Innate immunity is an evolutionarily ancient part of the host defense mechanisms: The same molecular modules are found in plants and animals, meaning that it arose before the split into these two kingdoms (2). Adaptive immunity is a relative newcomer on the evolutionary landscape. Because the mechanism of generating receptors in the adaptive immune system involves great variability and rearrangement of receptor gene segments, the adaptive immune system can provide specific recognition of foreign antigens, immunological memory of infection, 0732-0582/02/0407-0197$14.00

197

1 Feb 2002

10:13

198

AR

AR152-08.tex

JANEWAY

¥

AR152-08.SGM

LaTeX2e(2001/05/10)

P1: GJC

MEDZHITOV

Annu. Rev. Immunol. 2002.20:197-216. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

TABLE 1 Innate and adaptive immunity Property

Innate immune system

Adaptive immune system

Receptors

Fixed in genome Rearrangement is not necessary

Encoded in gene segments Rearrangement necessary

Distribution

Non-clonal All cells of a class identical

Clonal All cells of a class distinct

Recognition

Conserved molecular patterns (LPS, LTA, mannans, glycans)

Details of molecular structure (proteins, peptides, carbohydrates)

Self-Nonself discrimination

Perfect: selected over evolutionary time

Imperfect: selected in individual somatic cells

Action time

Immediate activation of effectors

Delayed activation of effectors

Response

Co-stimulatory molecules Cytokines (IL-1β, IL-6) Chemokines (IL-8)

Clonal expansion or anergy IL-2 Effector cytokines: (IL-4, IFNγ )

and pathogen-specific adaptor proteins. However, the adaptive immune response is also responsible for allergy, autoimmunity, and the rejection of tissue grafts. Innate immunity also lies behind most inflammatory responses; these are triggered in the first instance by macrophages, polymorphonuclear leukocytes, and mast cells through their innate immune receptors. What adaptive immunity adds to the underlying innate immune system is specific recognition of proteins, carbohydrates, lipids, nucleic acids, and pathogens, using the same activated, but not antigen-specific, effector cells generated by innate immune recognition. So the two systems are also linked in the use of the same effector cells (1). However, the real question is, how are they linked in the generation of an adaptive immune response? Unfortunately, defects in innate immunity though very rare are almost always lethal. They are rarely observed in a physician’s office, unlike defects in adaptive immunity, and only appeared once the wonder drug penicillin became available to treat infections. Therefore, we have relatively few patients surviving the lack of one or the other of their innate immune mechanisms, and thus we have relatively little data on the role of the innate immune system from such patients. In this article, we focus on how the innate immune system plays a role in discrimination between what we like to call infectious non-self and its obverse, which we refer to as noninfectious self. That is, we believe that the major decision to respond or not respond to a particular ligand is in the main decided by the genome-encoded innate immune system receptors. The positive and negative selection of developing lymphocytes plays a secondary but important role in this decision. We also should not forget recent evidence that a third type of cell could participate in the discrimination of self and nonself. These are called by various authors suppressor T cells (Ts) or regulatory T cells (Treg). But we argue that the main decision to respond or not to respond to a particular antigen is made by

1 Feb 2002

10:13

AR

AR152-08.tex

AR152-08.SGM

LaTeX2e(2001/05/10)

P1: GJC

Annu. Rev. Immunol. 2002.20:197-216. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

INNATE IMMUNE RECOGNITION

199

innate immune recognition receptors when they encounter pathogen-associated molecular patterns (PAMPs), such as LPS or bacterial CpG DNA. The innate immune system is made of many cells, such as those white blood cells that are not B lymphocytes or T lymphocytes of the adaptive immune system. It also can be considered to be a property of the skin and the epithelia that line our internal organs such as the gut and lungs. These tissues are normally populated with what are called commensal microbes, although when one gets a cut or, more seriously, a perforating ulcer, these commensals become pathogens. This is also seen in individuals who receive antibiotics, the main effect of which is to kill most or all microorganisms, with the result that there is frequently overgrowth of pathogenic microbes. This could in part be due to the loss of Escherichia coli, which produce potent antimicrobial peptides called colicins; in the absence of colicins, other, more dangerous pathogens may grow out and colonize the gut. There are many aspects to innate immunity that fall outside the purview of this article, but they are nevertheless important components of host defense. Among these are antimicrobial peptides produced by polymorphonuclear leukocytes in most vertebrate species and by the fat body in the fruit fly Drosophila melanogaster. The complement pathway can also be triggered by the mannose-binding lectin, an acute phase protein. That makes the control of these responses important, and we discuss them later in the article. Among the cells that bear innate immune or germline-encoded recognition receptors are macrophages, dendritic cells (DCs), mast cells, neutrophils, eosinophils, and the so-called NK cells. These cells can become activated during an inflammatory response, which is virtually always a sign of infection with a pathogenic microbe. Such cells rapidly differentiate into short-lived effector cells whose main role is to get rid of the infection; in this they mainly succeed without recourse to adaptive immunity. However, in certain cases, the innate immune system is unable to deal with the infection, and so activation of an adaptive immune response becomes necessary. In these cases, the innate immune system can instruct the adaptive immune system about the nature of the pathogenic challenge. It does so through the expression of costimulatory molecules, such as CD80 and CD86, on the surface of specialized antigen-presenting cells, the most important of which are the dendritic cells that guard against infection in virtually all tissues (1, 3, 4). Tissue dendritic cells also play a featured role in the initiation of tissue graft rejection. The genes for CD80 and CD86 are regulated by a transcription factor called NFκB, which can lead to rapid induction of their expression on the cell surface of dendritic cells and other antigen-presenting cells. The mechanism of NF-κB regulation is discussed in detail in this article, and it also has been reviewed extensively in this series (5). The T and B lymphocytes of the adaptive immune system have receptors that need to be assembled from gene segments. This allows great variability in adaptive immune recognition, but it cannot discern the nature of the pathogen infecting the body (6). Adaptive immunity also allows the unfortunate effects of autoimmune

1 Feb 2002

10:13

Annu. Rev. Immunol. 2002.20:197-216. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

200

AR

AR152-08.tex

JANEWAY

¥

AR152-08.SGM

LaTeX2e(2001/05/10)

P1: GJC

MEDZHITOV

disease, allergy, and allograft rejection. These latter are all a consequence of immune responses to nonpathogen antigens. They also appear to be a result of the random nature of receptor gene segment rearrangement. Nevertheless, in animals or plants with only an innate immune system at their disposal, there is no sign of any of these processes, despite numerous attempts to demonstrate such effects. Instead, what appears to be graft rejection is more readily explained by invoking cells like NK cells, which recognize the absence of self rather than the presence of nonself. Whether this is at the root of the mechanism is outside the scope of this article. However, invertebrates and plants lack the essential genes to make an adaptive immune response, and they also lack the associated tissue architecture to induce such a response. Therefore, the problems generated by having an adaptive immune system, as well as the benefits of having one, are found only in the descendents of the teleost fish, including ourselves. In recent times, the origin of the adaptive immune response has been uncovered. It turns out that the two recombinase-activating genes are encoded in a short stretch of DNA, in opposite orientations and lacking exons. This suggested an origin in a retroposon, as did the presence of the recognition signal sequences that lie 30 of all V gene segments and 50 of all J gene segments (7). This hypothesis was tested in vitro and shown to be true (8, 9). Other processes expand diversity tremendously, such as the generation of D gene segments in the first chain to rearrange, the nucleotide-adding enzyme TdT that inserts nucleotides in the junctions of V-D-J junctions, and somatic hypermutation. All of these processes are found exclusively in vertebrates and not in plants and invertebrates. Therefore, these organisms are wonderful tools for studying innate immunity, and the fact that the earth’s surface is covered with more species of invertebrates and plants than vertebrates speaks to the success of innate immune systems. This gives testimony to the necessity of having both an adaptive and an innate immune system; the vertebrates just added an antigen-specific mechanism for recognizing specific pathogens to a pre-existing system of non-antigen-specific innate immunity. Some of the advantages of having an adaptive immune system are the ability to remember or adapt to an infectious agent, but this memory is confined to an individual. Apart from the trans-placental transfer of antibody from mother to fetus, there is little carryover of this memory from one generation to the next. This may be good because many have speculated that adaptive immunity toward pathogenic microbes is an initiating signal for autoimmunity (10). One can think in very simple terms about the virtues of a nonclonal system of host defense. First, it serves to make adaptive immunity more useful, in part by delaying the need for an adaptive immune response by the three to five days that it takes to generate the clonal expansion and differentiation to effector lymphocytes. Second, it serves to alert the clonal, adaptive immune system that it is under attack by a pathogenic microorganism. The nonclonal system also cannot mediate all the bad effects of adaptive immunity because it involves rapid activation of effector cells that, if they were directed against self tissues, would be lethal to the host and

1 Feb 2002

10:13

AR

AR152-08.tex

AR152-08.SGM

LaTeX2e(2001/05/10)

P1: GJC

Annu. Rev. Immunol. 2002.20:197-216. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

INNATE IMMUNE RECOGNITION

201

thus expunged by evolutionary processes (6). The virtue of having both innate and adaptive systems of recognition is that the interplay of these two distinct systems allows the discrimination of an infectious attack on the host from noninfectious self. The virtue of having clonally distributed receptors is also obvious, in that they allow the recognition of particular features of pathogens. Such receptors also remember having seen pathogens before, and so they mount a stronger and more specific response on re-encounter with the same microbe. Those organisms, all of which are vertebrates, that have both innate and adaptive immune systems gain a tremendous benefit in longevity through stimulation of having both systems. This may be of particular importance in vaccination, which was initiated a bit over 200 years ago by Edward Jenner. Recently, it was reported that the vaccinia virus introduced by Jenner had genetic elements that could inhibit an intracellular domain found in innate immune recognition receptors (11). This finding, just made, says that the vaccinia can grow for a while in a nonvaccinated individual because it can shut off the innate immune response. This property is undoubtedly of importance to the virus in infecting cows (and milkmaids), but it can also allow vaccinia to be a very potent vaccine that leads to resistance to small pox of variola, with which it cross-reacts. Thus, the virtues of having an innate immune system of pathogen recognition lie not only in the delaying tactics of inflammation upon infection, but also in the activation of the adaptive immune system only when the body is under attack by a specific pathogen. This system works to allow long life in most vertebrates by controlling the expression of the cell surface costimulatory molecules, and by inducing secretion of appropriate cytokines and chemokines that direct the lymphocytes of the adaptive immune system to their appropriate locations. Together these function to give optimal host defense.

PATTERN RECOGNITION RECEPTORS The innate immune system uses a variety of pattern recognition receptors that can be expressed on the cell surface, in intracellular compartments, or secreted into the bloodstream and tissue fluids (12). The principal functions of pattern recognition receptors include opsonization, activation of complement and coagulation cascades, phagocytosis, activation of proinflammatory signaling pathways, and induction of apoptosis. Mannan-binding lectin (MBL), C-reactive protein (CRP), and serum amyloid protein (SAP) are secreted pattern recognition molecules produced by the liver during the acute phase response at the early stages of infection (13–15). CRP and SAP are members of the pentraxin family, and both can function as opsonins upon binding to phosphorylcholine on bacterial surfaces (13, 14). CRP and SAP can also bind to C1q and thus activate the classical complement pathway (16). MBL is a member of the collectin family, which also includes pulmonary surfactant proteins A and D (17, 18). The collectins are characterized by the presence of a collagenous

1 Feb 2002

10:13

Annu. Rev. Immunol. 2002.20:197-216. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

202

AR

AR152-08.tex

JANEWAY

¥

AR152-08.SGM

LaTeX2e(2001/05/10)

P1: GJC

MEDZHITOV

region and a C-type lectin (CTL) domain; typically they form oligomeric receptors (17, 18). MBL binds specifically to terminal mannose residues, which are abundant on the surface of many microorganisms, and associates with MBL-associated serine proteases (MASP). MASP1 and MASP2 are activated by MBL and initiate the lectin pathway of complement by cleaving C2 and C4 proteins (15). Several cell surface receptors expressed on macrophages function as pattern recognition receptors that mediate phagocytosis of microorganisms. Macrophage mannose receptor (MMR) is a member of the C-type lectin family and is closely related to DEC205, a receptor expressed on dendritic cells. MMR interacts with a variety of gram-positive and gram-negative bacteria and fungal pathogens (15). The main function of the MMR is thought to be phagocytosis of microbial pathogens, and their delivery into the lysosomal compartment where they are destroyed by lysosomal enzymes (15). The function and ligand specificity of DEC205 has not yet been characterized, but its similarity to MMR and its expression on dendritic cells suggest that it may also function as a phagocytic receptor. Macrophage scavenger receptor (MSR) is another phagocytic pattern recognition receptor expressed on macrophages. MSR belongs to the scavenger receptor type A (SR-A) family and has an unusually broad specificity to a variety of polyanionic ligands, including double-stranded RNA (dsRNA), LPS, and LTA (19). MSR protects against endotoxic shock by scavenging LPS and has a role in host defense, as demonstrated by increased susceptibility of MSR-deficient mice to Listeria monocytogenes, herpes simplex virus, and malaria infections (20, 21). In addition to recognition of microbial PAMPs, MSR also plays a role in lipid homeostasis by binding and endocytosing acetylated low-density lipoproteins (22). Another SR-A family member, MARCO, is a macrophage receptor that binds to bacterial cell walls and LPS, and it also mediates phagocytosis of bacterial pathogens (23).

INTRACELLULAR RECOGNITION SYSTEMS Viruses and some bacterial pathogens can gain access to the intracellular compartments, such as the cytosol. Several pattern recognition receptors are expressed in the cytosol where they detect these intracellular pathogens and induce responses that block their replication. The protein kinase PKR is activated upon binding to dsRNA, which is produced during viral infection (24). Activated PKR phosphorylates and inactivates the translation initiation factor eIF2α, which results in a block of viral and cellular protein synthesis (24). In addition, PKR activates NF-κB and MAP kinase signaling pathways, which leads to the induction of the antiviral type-I IFN genes (25). PKR also inhibits viral spread by inducing apoptosis in infected cells (25). Another antiviral pathway activated by dsRNA is the 20 -50 -oligoadenylate synthase (OAS)/RNaseL pathway (26). OAS is activated upon binding to viral dsRNA and produces an unusual nucleotide second messenger - 20 -50 oligoadenylate. These oligonucleotides then activate the RNaseL, which destroys both viral and cellular

1 Feb 2002

10:13

AR

AR152-08.tex

AR152-08.SGM

LaTeX2e(2001/05/10)

P1: GJC

Annu. Rev. Immunol. 2002.20:197-216. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

INNATE IMMUNE RECOGNITION

203

RNAs. The antiviral effect of this pathway is therefore due to cleavage of viral RNA and induction of apoptosis in infected cells due to cleavage of cellular RNA and block of protein synthesis (26). Another group of proteins likely involved in intracellular pattern recognition is the family of NOD proteins. NOD proteins contain an N-terminal CARD domain, a nucleotide binding domain (NBD), and a C-terminal leucine-rich repeat (LRR) region (27, 28). This domain arrangement is characteristic of the NB-ARC family, which, in addition to the NODs, includes the mammalian apoptosis regulator APAF-1 and a class of plant resistance genes (R genes). R genes in plants can detect microbial infection and induce the hypersensitivity response—a major effector response that blocks pathogen replication and spread (29). The CARD domains of NOD1 and NOD2 associate with a protein kinase, RIP2, which in turn activates NF-κB and MAP kinase signaling pathways (27, 28, 30). The full range of ligands recognized by NOD proteins is currently unknown, but both NOD1 and NOD2 are reported to activate NF-κB in response to LPS, presumably through binding to their LRR regions (30, 31). It is interesting that mutations in the nod2 gene cause a predisposition to Crohn’s disease—a chronic inflammatory disorder of the gut (32, 33).

DROSOPHILA TOLL The first member of the Toll family, Drosophila Toll, was discovered as one of 12 maternal effect genes that function in a pathway required for dorso-ventral axis formation in fly embryos (34, 35). Other genes in this pathway encode a Toll ligand, Spatzle, an adapter protein, Tube, a protein kinase, Pelle, an NF-κB family transcription factor, Dorsal, and a Dorsal inhibitor, Cactus, which is a homologue of mammalian IκB. Spatzle is secreted as a precursor polypeptide and requires proteolytic cleavage by serine proteases for activation. This cleavage is controlled by a protease cascade that includes four serine proteases: gastrulation defective, easter, snake, and nudel (34, 36). Soon after Drosophila Toll and the human IL-1R were identified, it became apparent that they had possible functional similarities. In addition to the presence of homologous cytoplasmic TIR domains, both receptors could induce NF-κB activation and could signal through homologous protein kinases—Pelle and IRAK (34, 36). The analysis of the promoter regions of the genes encoding antimicrobial peptides in Drosophila revealed consensus NF-κB binding sites (37). Since these peptides are rapidly induced in flies in response to infection, these observations suggested a possible involvement of the Toll pathway in Drosophila immunity. Indeed, analysis of Drosophila strains carrying loss-of-function mutations in the Toll gene demonstrated a striking defect in immune responses: These flies were highly susceptible to fungal infection but had normal responses to gram-negative bacterial infection (38). The systemic immune response in Drosophila is mediated by a battery of antimicrobial peptides produced largely by the fat body, an insect organ analogous to

1 Feb 2002

10:13

Annu. Rev. Immunol. 2002.20:197-216. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

204

AR

AR152-08.tex

JANEWAY

¥

AR152-08.SGM

LaTeX2e(2001/05/10)

P1: GJC

MEDZHITOV

the mammalian liver. These peptides lyse microorganisms by forming pores in their cell walls. Functionally, the antimicrobial peptides fall into three classes depending on the pathogen specificity of their lytic activity (2). Thus, Drosomycin is a major antifungal peptide, whereas Diptericin is active against gram-negative bacteria, and Defensin works against gram-positive bacteria (2). Interestingly, infection of Drosophila with different classes of pathogens leads to preferential induction of the appropriate group of antimicrobial peptides (39). For example, fungal infection results in the induction of Drosomycin, but not Diptericin (39). Mutation of the Toll gene blocks the induction of Drosomycin in response to fungal infection but does not affect significantly the induction of Diptericin in response to gram-negative infection. Importantly, mutations in Spatzle, Tube, Pelle, and Cactus genes also specifically affect the resistance of Drosophila to fungal pathogens (38). Dorsal, which is activated by the Toll pathway during dorso-ventral axis formation, does not appear to play a role in the systemic immune response in adult flies. Instead, another NF-κB family member—Dif (drosophila immunity factor)—is required for the induction of Drosomycin by Toll (40–42). Additionally, Spatzle is required for the activation of Toll by fungal pathogens; however, the serine protease cascade that generates active Spatzle during development is not involved in the immune response (38). Therefore, a different protease cascade must regulate its processing. This putative protease cascade is presumably triggered by a PRR(s) specific for fungal PAMPs, such as mannan. Further support for this hypothesis came from the analysis of necrotic mutants (43). Necrotic encodes a serine protease inhibitor of the serpin family. Mutations in this gene result in the spontaneous activation of the Toll pathway and constitutive induction of the Drosomycin gene (43). These results suggest that in Drosophila, the pattern recognition event occurs upstream of Toll and triggers a protease cascade, much as complement is activated by the lectin pathway in mammals. Interestingly, the Toll pathway can also be activated in response to gram-positive infection, suggesting that multiple pattern recognition molecules may function upstream of the protease cascade that controls cleavage of Spatzle (44). The Drosophila response to gram-negative bacterial infection is controlled by a distinct pathway, which was defined by the mutation in the imd (immune deficient) gene (45). Imd mutants have a profound defect in resistance to gram-negative bacterial pathogens, while remaining essentially normal with regard to fungal and gram-positive infection (45). Genetic analyses led to the identification of four additional genes that function in the Imd pathway: Dredd, dIKK-γ , dIKK-β, and Relish (46–51). Mutations in any of these genes yield phenotypes very similar to the Imd mutants, that is, susceptibility to gram-negative bacterial infection due to impaired induction of antibacterial peptides such as Diptericin. Dredd is a Drosophila caspase previously implicated in the control of apoptosis during fly development (52). dIKK-γ and dIKK-β are Drosophila homologues of human IKK-γ (also known as NEMO) and IKK-β. In human cells, IKK-β and NEMO are essential regulators of NF-κB activation (53). Relish is a Drosophila homologue of the mammalian Rel/NF-κB family members, p100 and p105 (54).

1 Feb 2002

10:13

AR

AR152-08.tex

AR152-08.SGM

LaTeX2e(2001/05/10)

P1: GJC

Annu. Rev. Immunol. 2002.20:197-216. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

INNATE IMMUNE RECOGNITION

205

One major question in Drosophila immunity that remains unresolved is the identity of the receptor that controls activation of the imd pathway in response to gramnegative bacterial infection. Since there are nine Toll-like receptors in Drosophila, it is possible that one of them may be responsible for the activation of the imd pathway (55). A mutation in 18-Wheeler, another Toll family member, affected expression of the antibacterial peptide, Attacin (56). However 18-Wheeler does not appear to function in the imd pathway (44). Moreover, none of the Drosophila Tolls could induce activation of the Diptericin promoter in Drosophila cell lines, and only Toll and Toll-5 were able to activate Drosomycin (55). Alternatively, a receptor unrelated to Toll may control the imd pathway and function as a sensor for gram-negative PAMPs such as LPS.

TOLL-LIKE RECEPTORS IN MAMMALIAN IMMUNITY Ten TLRs have been described to date in humans and mice. They differ from each other in ligand specificities, expression patterns, and presumably in the target genes they can induce. No developmental function has been ascribed to mammalian TLRs so far. Several TLRs are involved in the recognition of a variety of PAMPs. The exact mechanism of recognition has not yet been determined for any of them and remains an important avenue of future research.

LIGAND RECOGNITION BY TLRS TLR4 The first indication that mammalian TLRs may function as pattern recognition receptors came with the description of a human homologue of Drosophila Toll, now known as TLR4 (57). A constitutively active form of this receptor induced the expression of inflammatory cytokines and the costimulatory molecule B7 in the monocytic cell line THP-I (57). Subsequently, positional cloning analysis of the LPS-nonresponsive mouse strain, C3H/HeJ, showed that a point mutation in the TIR domain of TLR4 was responsible for the defect in LPS signal transduction (58, 59). Another mouse strain, B10.ScCR, did not respond to LPS and turned out to lack the genomic region that contains the entire tlr4 gene (58, 59). Finally, mice with a targeted deletion of the TLR4 gene were unresponsive to LPS (60). Together, these studies demonstrated the essential role for TLR4 in recognition of a major component of gram-negative bacteria. TLR4, however, is not the sole receptor involved in LPS recognition. Transport of LPS molecules in the serum is mediated by LPS-binding protein (LBP) (61). At the plasma membrane, LBP is thought to transfer LPS monomers to CD14, a GPI-linked cell surface protein (61). Exactly how CD14 facilitates recognition of LPS by TLR4 is not clear, but its critical role is underscored by the LPShyporesponsive phenotype of CD14-deficient mice (62, 63). Finally, a small protein called MD-2 is also a component of the LPS-recognition complex (64). MD-2 lacks

1 Feb 2002

10:13

Annu. Rev. Immunol. 2002.20:197-216. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

206

AR

AR152-08.tex

JANEWAY

¥

AR152-08.SGM

LaTeX2e(2001/05/10)

P1: GJC

MEDZHITOV

a transmembrane anchor but is associated with the extracellular region of TLR4 (64). MD-2 is required for cellular responsiveness to LPS, as demonstrated by both transfection studies and an analysis of a CHO cell line with a mutated MD-2 gene (64, 65). Although the cell-surface events that confer LPS recognition have not been unambiguously determined, most of the available evidence indicates that a complex of TLR4/MD-2/CD14 directly binds LPS (66–68). How other members of the mammalian TLR family recognize their cognate ligands is still an enigma, and one that waxes in complexity as more and more ligands are identified for some individual members of the TLR family. Interestingly, B cells express on their cell surface a receptor called RP105 that is also involved in LPS recognition (69). RP105 is related to TLR4 in its extracellular domain, which likewise consists of leucine-rich repeats, but it lacks the intracellular TIR domain and instead has a short cytosolic tail that contains a tyrosine phosphorylation motif (69). RP105 is associated via its ectodomain with MD-1, a protein related to MD-2 that is required for RP-105 function (70, 71). In response to cross-linking or LPS stimulation, RP105 activates src kinases, including lyn (72). Since B cell responses to LPS are completely dependent on TLR4, the exact mechanism of LPS recognition is unclear, but it presumably involves cooperation between RP105 and TLR4 (72). In addition to LPS, TLR4 has been implicated in the recognition of liptocheic acid (LTA), the heat shock protein hsp60, and the fusion protein of the respiratory syncytial virus (73–76). The physiological relevance of some of these putative TLR4 ligands remains to be demonstrated. However, it is clear that the original paradigm suggested by the example of Drosophila Tolls, in which different Tolls discriminate between classes of pathogens, is not applicable to mammalian TLRs.

TLR2 Of the mammalian TLRs, and perhaps of all PRRs, TLR2 recognizes the largest number of ligands. The list includes peptidoglycan (73, 77), bacterial lipoproteins (78–80), a phenol soluble factor from Staphylococcus epidermidis (81), LPS from Prophyromonas gingivitis (82) and Leptospira interrogans (which differs in structure from the LPS of gram-negative bacteria) (83), glycosylphosphotidylinositol lipid from Trypanosoma cruzi (84), and zymosan, a component of yeast cell walls (85). TLR2 does not recognize these PAMPs independently, but functions by forming heterodimers with either TLR1 or TLR6 (86, 87). A likely consequence of this cooperation is an increased repertoire of ligand specificities. Further studies are needed to determine whether heterodimerization is necessary for ligand recognition by any other TLR. No other TLR pairs have yet been identified, and some of the TLRs (such as TLR4 and TLR5) most likely function as homodimers (86).

TLR5 TLR5 recognizes flagellin, the protein subunits that make up bacterial flagella (88). Unlike most other PAMPs, flagellin is a protein and does not contain any obvious features to flag it as nonself or pathogen-associated. Nevertheless, flagellin

1 Feb 2002

10:13

AR

AR152-08.tex

AR152-08.SGM

LaTeX2e(2001/05/10)

P1: GJC

Annu. Rev. Immunol. 2002.20:197-216. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

INNATE IMMUNE RECOGNITION

207

is extremely conserved at the N- and C-terminal ends that form its hydrophobic core and most likely is recognized by TLR5 in this region. It is probable that structural constraints have prevented mutations in these conserved regions and hence generation of escape mutants. Moreover, flagella, like other PAMPs, are essential for viability, and a mutation that compromises flagellin function would have deleterious consequences for the bacteria. Flagellin from S. adelaide played an important role in early studies of immunity. It was a highly potent antigen that turned out to be thymus independent. It played a major role in establishing the clonal selection hypothesis of MacFarlene Burnet. The fact that this antigen was used intensively in Australia may account for most of the success of Australian immunology in the 1960s and 1970s; it drew numerous people to work at the Walter and Elisa Hall Institute.

TLR9 Unmethylated CpG DNA was long known for its immunostimulatory effects (89, 90), and we now know that TLR9 recognizes unmethylated CpG motifs present in bacterial DNA (91). The logic of this recognition is that most of the mammalian genome is methylated, while bacteria lack CpG methylation enzymes (90). One enigma concerning this PAMP is in its accessibility for recognition by TLR9; bacterial DNA, after all, should be neatly packaged in the bacteria and rarely, if ever, exposed for recognition at the bacterial cell surface. However, stimulation by CpG DNA can be inhibited by drugs that block its uptake, and therefore, TLR9 likely recognizes its ligand intracellularly, perhaps in endosomes or lysosomes, presumably following bacterial lysis (89, 92). It is worth noting here that although all the TLRs are assumed to reside at the cell surface, some of the TLRs (including TLR9) may in fact localize intracellularly. Furthermore, in some cases, cell-surface ligand binding may also be coupled to uptake, such that the TLR undergoes stimulusdependent internalization, in the process delivering its cargo to an intracellular compartment. Thus, TLR2, for example, is recruited to macrophage phagosomes upon stimulation with zymosan (85).

TOLL SIGNALING PATHWAYS Upon recognition of their cognate ligands, TLRs induce the expression of a variety of host defense genes. These include inflammatory cytokines and chemokines, antimicrobial peptides, costimulatory molecules, MHC molecules, and other effectors necessary to arm the host cell against the invading pathogen. TLRs accomplish this by activating an intracellular signaling pathway conserved from Drosophila to mammals. Furthermore, this pathway is remarkably similar to the one activated by the IL-1R (which also has a cytosolic TIR domain); indeed, identical molecules comprise the two signaling cascades (93), and until very recently we knew of no signaling components unique to one or the other pathway. Upon ligation of TLR4 (and IL-1R), the adapter MyD88 is recruited to the receptor complex (94, 95). MyD88 has a C-terminal TIR domain that mediates

1 Feb 2002

10:13

Annu. Rev. Immunol. 2002.20:197-216. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

208

AR

AR152-08.tex

JANEWAY

¥

AR152-08.SGM

LaTeX2e(2001/05/10)

P1: GJC

MEDZHITOV

its homophilic interaction with the receptor and an N-terminal death domain that engages the death domain of its downstream target IRAK (96). IRAK may be recruited to the receptor via Tollip, an adapter that contains a C2 domain (97). Upon association with MyD88, IRAK, a serine threonine kinase, undergoes autophosphorylation. The RING-finger containing adapter TRAF6 is also part of this activated signaling complex (98). One study suggested that TRAF6 functions as an E3 ligase to ubiquitinate an as-yet-unidentified target that is necessary for TLRand IL-1R-mediated IκB kinase β (IKK-β activation) (99). Activated IKK phosphorylates and targets for degradation the NF-κB inhibitor IκB, thereby freeing NF-κB to translocate into the nucleus and turn on transcription of target genes (5). Although all TLRs signal through the conserved signaling cascade described above, the complexity of the TLR-induced cellular responses indicates that there must be additional regulatory mechanisms and signaling pathways downstream of TLRs. One example is provided by the existence of a Rac1-PI3K-AKT pathway activated by TLR2. This pathway leads ultimately to phosphorylation of NF-κB and is necessary for NF-κB transactivation activity (100). It will be important to test whether other TLRs also activate a similar pathway that enhances NF-κB transactivation potential. Because Rac1, PI3K, and AKT regulate diverse cellular functions in other pathways, this study also raises the interesting possibility of links connecting the TLR pathway to other signaling pathways. Although analyses of knockout mice have confirmed that TLRs and IL-1Rs share in common many of the same signaling components, the members of the two families can induce distinct targets. Even different TLR members activate distinct albeit overlapping sets of target genes. Undoubtedly, there must be mechanisms that enable TLRs to achieve specificity in activation of cellular responses. The first indication of such a mechanism came with the analyses of MyD88-deficient mice (101–103). As expected, these mice were unable to activate NF-κB and MAP kinases, or to upregulate surface expression of MHC and costimulatory molecules in response to IL-1 and many TLR ligands, including peptidoglycan and unmethylated CpG motifs (80, 101–104). Surprisingly, however, the TLR4 ligand LPS could still activate NF-κB and MAP kinases (albeit with delayed kinetics) in the absence of MyD88 (101). Moreover, LPS-stimulated MyD88deficient dendritic cells retain the ability to upregulate costimulatory and MHC molecules (102). Therefore, although MyD88 is required for all signaling events downstream of some TLRs, such as TLR2 and TLR9, MyD88 is clearly dispensable for some TLR4-induced signals. However, other inflammatory responses appear to be completely dependent on MyD88 regardless of the stimulus; MyD88-deficient mice do not produce the cytokine IL-12 in response to any of the tested PAMPs, including LPS and CpG DNA (102, 104). These studies indicate that whereas signaling downstream of some TLRs (such as TLR9) and IL-1R is completely dependent on MyD88, TLR4 can activate two pathways, a MyD88-dependent pathway similar to that activated by other TLRs and IL-1R and a MyD88-independent pathway. Furthermore, the existence of a MyD88-independent pathway suggested that TLR4 may transduce some signals through a distinct adapter protein. This hypothesis was borne out by the

1 Feb 2002

10:13

AR

AR152-08.tex

AR152-08.SGM

LaTeX2e(2001/05/10)

P1: GJC

Annu. Rev. Immunol. 2002.20:197-216. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

INNATE IMMUNE RECOGNITION

209

identification of the novel TIR domain-containing adapter protein, TIRAP (105). TIRAP contains a small N-terminal region of unknown function and a C-terminal TIR domain that mediates its interaction with TLR4. A dominant negative mutant of TIRAP specifically inhibits TLR4- but not IL-1R- or TLR9-induced NF-κB activation, indicating a specificity of TIRAP for the TLR4 pathway. TIRAP therefore represents the first identified molecule responsible for at least some of the known differences in signaling between TLR4 and other TLRs, as well as between TLRs and the IL-1Rs (105). Although many of the other components and targets of this TIRAP-regulated pathway remain to be identified, one downstream target of the TIRAP pathway is PKR, the interferon-regulated, dsRNA-activated protein kinase (105). Indeed, PKR is also a component of the MyD88-dependent pathway, as stimulation of wild-type (but not MyD88-deficient) macrophages with CpG results in PKR activation. Therefore, PKR, previously identified as a component of antiviral defense and stress responses, also functions in TLR signaling pathways (105).

Toll and Control of Adaptive Immunity Dendritic cells are pivotally positioned at the interface of innate and adaptive immunity (4). Immature dendritic cells reside in the peripheral tissues, where they actively sample their environment by endocytosis and macropinocytosis. Upon encountering a pathogen, they undergo a developmental program called dendritic cell maturation, which includes induction of costimulatory activity, antigen processing, increased MHC molecule expression, and migration to the lymph node, where they can prime na¨ıve antigen-specific T cells (4). In this way activation of the adaptive immune system occurs only upon pathogen recognition by dendritic cells. Pathogen recognition, of course, is mediated by TLRs on the surface of dendritic cells; not surprisingly, these cells express high levels of most members of the TLR family. Analysis of MyD88-deficient mice demonstrated the critical role of TLRs in DC maturation and induction of adaptive immune responses. As mentioned above, stimulation of MyD88-deficient DCs with all tested PAMPs except LPS does not result in DC maturation (101, 106). Concomitant with the block in surface expression of costimulatory and MHC molecules, these DCs, not surprisingly, cannot prime antigen-specific na¨ıve T cells in vitro (106). Furthermore, when MyD88-deficient mice are immunized with ovalbumin in complete Freund’s adjuvant (CFA), no ovalbumin-specific T cell responses develop. In addition, these mice also fail to produce IFN-γ and ovalbumin-specific IgG2a antibodies (106). These defects are due at least in part to the inability of MyD88-deficient dendritic cells to make IL-12. Remarkably, however, B cells in MyD88-deficient mice make normal amounts of antigen-specific IgG1 and IgE, while T cells produce higher levels of IL-13 upon restimulation than their counterparts from wild-type mice (106). Clearly, then, these mice appear to have a selective defect for mounting Th1 but not Th2 responses, and TLRs seem to control induction of only Th1-type inflammation (106). Indeed, all the known PAMPs are derived from either prokaryotic, fungal, viral, or protozoan pathogens, which are conventionally targets of Th1 responses. TLRs are

1 Feb 2002

10:13

Annu. Rev. Immunol. 2002.20:197-216. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

210

AR

AR152-08.tex

JANEWAY

¥

AR152-08.SGM

LaTeX2e(2001/05/10)

P1: GJC

MEDZHITOV

essential for recognizing these categories of pathogens and for mounting adaptive immune responses, which are generally Th1-type responses that are appropriate for their elimination. Th2 responses, on the other hand, combat multicellular eukaryotic parasites, which probably are recognized not by TLRs but by a distinct set of pattern-recognition receptors. An interesting implication is that allergens, which also elicit Th2 responses, probably also activate the immune system through TLR-independent mechanisms. Finally, the in vivo studies with the MyD88-deficient mice also demonstrate that adjuvants, an essential component of most vaccines, exert their immunostimulatory effects through activation of TLRs (106). The active ingredient of CFA, for example, is mycobacterial lysate, a TLR ligand (107, 108). Therefore, adjuvants function by stimulating TLRs expressed on dendritic cells, which in turn leads to dendritic cell maturation and the induction of antigen-specific adaptive immune responses (1, 106).

CONCLUSION Innate immune recognition is very complex, as it has to be to protect the host against a highly diverse microbial world. But it seems to be in essence much simpler than the adaptive immune response, which operates by recognizing fine details of pathogenic microorganisms. In order to respond, na¨ıve T cells need to recognize both the antigen bound to self-MHC ligands and a molecule of CD80 and/or CD86. These two proteins have to be expressed on the same antigen-presenting cell. The costimulatory molecules are induced by Toll-like receptors. There are ten TLRs in humans and mice, and they or their homologues are found in all multicellular organisms. Once a pathogen is recognized, the host antigen-presenting cell expressed on its surface costimulatory molecules and in the cytosol proinflammatory cytokines and chemokines. These molecules, together, can both attract na¨ıve T cells through the secretions of chemokines and activate na¨ıve T cells to respond to specific antigens of the pathogen. These antigens are displayed on the same cell surface as the induced costimulatory molecules, providing both an antigen-specific stimulus and the required costimulatory molecules to activate na¨ıve, antigen-specific T cells. Once T cells are activated, the adaptive immune response takes over, and the pathogen is engulfed by a phagocyte and destroyed. Visit the Annual Reviews home page at www.annualreviews.org

LITERATURE CITED 1. Janeway CA Jr. 1989. Approaching the asymptote? Evolution and revolution in immunology. Cold Spring Harbor Symp. Quant. Biol. 54:1–13 2. Hoffmann JA, Kafatos FC, Janeway CA,

Ezekowitz RA. 1999. Phylogenetic perspectives in innate immunity. Science 284:1313–18 3. Fearon DT, Locksley RM. 1996. The instructive role of innate immunity in the

1 Feb 2002

10:13

AR

AR152-08.tex

AR152-08.SGM

LaTeX2e(2001/05/10)

P1: GJC

INNATE IMMUNE RECOGNITION

4.

5.

Annu. Rev. Immunol. 2002.20:197-216. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

6.

7.

8.

9.

10.

11.

12.

13.

14.

acquired immune response. Science 272: 50–53 Banchereau J, Steinman RM. 1998. Dendritic cells and the control of immunity. Nature 392:245–52 Ghosh S, May MJ, Kopp EB. 1998. NF-kappa B and Rel proteins: evolutionarily conserved mediators of immune responses. Annu. Rev. Immunol. 16:225–60 Janeway CA Jr. 1992. The immune system evolved to discriminate infectious nonself from noninfectious self. Immunol. Today 13:11–16 Thompson CB. 1995. New insights into V(D)J recombination and its role in the evolution of the immune system. Immunity 3:531–39 Agrawal A, Eastman QM, Schatz DG. 1998. Transposition mediated by RAG1 and RAG2 and its implications for the evolution of the immune system. Nature 394:744–51 Hiom K, Melek M, Gellert M. 1998. DNA transposition by the RAG1 and RAG2 proteins: a possible source of oncogenic translocations. Cell 94:463–70 Oldstone MB, Nerenberg M, Southern P, Price J, Lewicki H. 1991. Virus infection triggers insulin-dependent diabetes mellitus in a transgenic model: role of anti-self (virus) immune response. Cell 65:319–31 Bowie A, Kiss-Toth E, Symons JA, Smith GL, Dower SK, O’Neill LA. 2000. A46R and A52R from vaccinia virus are antagonists of host IL-1 and toll-like receptor signaling. Proc. Natl. Acad. Sci. USA 97:10162–67 Medzhitov R, Janeway CA Jr. 1997. Innate immunity: impact on the adaptive immune response. Curr. Opin. Immunol. 9:4–9 Schwalbe RA, Dahlback B, Coe JE, Nelsestuen GL. 1992. Pentraxin family of proteins interact specifically with phosphorylcholine and/or phosphorylethanolamine. Biochemistry 31:4907–15 Gewurz H, Mold C, Siegel J, Fiedel B. 1982. C-reactive protein and the acute

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

211

phase response. Adv. Intern. Med. 27:345– 72 Fraser IP, Koziel H, Ezekowitz RA. 1998. The serum mannose-binding protein and the macrophage mannose receptor are pattern recognition molecules that link innate and adaptive immunity. Semin. Immunol. 10:363–72 Agrawal A, Shrive AK, Greenhough TJ, Volanakis JE. 2001. Topology and structure of the C1q-binding site on C-reactive protein. J. Immunol. 166:3998–4004 Holmskov UL. 2000. Collectins and collectin receptors in innate immunity. APMIS Suppl. 100:1–59 Epstein J, Eichbaum Q, Sheriff S, Ezekowitz RA. 1996. The collectins in innate immunity. Curr. Opin. Immunol. 8:29–35 Pearson AM. 1996. Scavenger receptors in innate immunity. Curr. Opin. Immunol. 8:20–28 Thomas CA, Li Y, Kodama T, Suzuki H, Silverstein SC, El Khoury J. 2000. Protection from lethal gram-positive infection by macrophage scavenger receptordependent phagocytosis. J. Exp. Med. 191:147–56 Suzuki H, Kurihara Y, Takeya M, Kamada N, Kataoka M, Jishage K, Ueda O, Sakaguchi H, Higashi T, Suzuki T, et al. 1997. A role for macrophage scavenger receptors in atherosclerosis and susceptibility to infection. Nature 386:292– 96 Krieger M, Herz J. 1994. Structures and functions of multiligand lipoprotein receptors: macrophage scavenger receptors and LDL receptor-related protein (LRP). Annu. Rev. Biochem. 63:601–37 Elomaa O, Kangas M, Sahlberg C, Tuukkanen J, Sormunen R, Liakka A, Thesleff I, Kraal G, Tryggvason K. 1995. Cloning of a novel bacteria-binding receptor structurally related to scavenger receptors and expressed in a subset of macrophages. Cell 80:603–9 Clemens MJ, Elia A. 1997. The doublestranded RNA-dependent protein kinase

1 Feb 2002

10:13

212

25.

26.

Annu. Rev. Immunol. 2002.20:197-216. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

27.

28.

29.

30.

31.

32.

33.

AR

AR152-08.tex

JANEWAY

¥

AR152-08.SGM

LaTeX2e(2001/05/10)

P1: GJC

MEDZHITOV

PKR: structure and function. J. Interferon Cytokine Res. 17:503–24 Williams BR. 1999. PKR; a sentinel kinase for cellular stress. Oncogene 18: 6112–20 Kumar M, Carmichael GG. 1998. Antisense RNA: function and fate of duplex RNA in cells of higher eukaryotes. Microbiol. Mol. Biol. Rev. 62:1415–34 Bertin J, Nir WJ, Fischer CM, Tayber OV, Errada PR, Grant JR, Keilty JJ, Gosselin ML, Robison KE, Wong GH, Glucksmann MA, DiStefano PS. 1999. Human CARD4 protein is a novel CED-4/Apaf1 cell death family member that activates NF-kappaB. J. Biol. Chem. 274:12955– 58 Inohara N, Koseki T, del Peso L, Hu Y, Yee C, Chen S, Carrio R, Merino J, Liu D, Ni J, Nunez G. 1999. Nod1, an Apaf1-like activator of caspase-9 and nuclear factor-kappaB. J. Biol. Chem. 274:14560– 67 Hammond-Kosack K, Jones J. 1997. Plant disease resistance genes. Annu. Rev. Plant Physiol. Plant Mol. Biol. 48:575– 607 Ogura Y, Inohara N, Benito A, Chen FF, Yamaoka S, Nunez G. 2001. Nod2, a Nod1/Apaf-1 family member that is restricted to monocytes and activates NFkappaB. J. Biol. Chem. 276:4812–18 Inohara N, Oura Y, Chen FF, Muto A, Nunez G. 2001. Human Nod1 confers responsiveness to bacterial lipopolysaccharides. J. Biol. Chem. 276:2551–54 Hugot JP, Chamaillard M, Zouali H, Lesage S, Cezard JP, Belaiche J, Almer S, Tysk C, O’Morain CA, Gassull M, Binder V, Finkel Y, Cortot A, Modigliani R, Laurent-Puig P, Gower-Rousseau C, Macry J, Colombel JF, Sahbatou M, Thomas G. 2001. Association of NOD2 leucine-rich repeat variants with susceptibility to Crohn’s disease. Nature 411:599– 603 Ogura Y, Bonen DK, Inohara N, Nicolae DL, Chen FF, Ramos R, Britton H,

34.

35.

36.

37.

38.

39.

40.

41.

42.

Moran T, Karaliuskas R, Duerr RH, Achkar JP, Brant SR, Bayless TM, Kirschner BS, Hanauer SB, Nunez G, Cho JA. 2001. A frameshift mutation in NOD2 associated with susceptibility to Crohn’s disease. Nature 411:603–6 Belvin MP, Anderson KV. 1996. A conserved signaling pathway: the Drosophila tolldorsal pathway. Annu. Rev. Cell Dev. Biol. 12:393–416 Hashimoto C, Hudson KL, Anderson KV. 1988. The Toll gene of Drosophila, required for dorsal-ventral embryonic polarity, appears to encode a transmembrane protein. Cell 52:269–79 Anderson KV. 2000. Toll signaling pathways in the innate immune response. Curr. Opin. Immunol. 12:13–19 Engstrom Y, Kadalayil L, Sun SC, Samakovlis C, Hultmark D, Faye I. 1993. kappa B-like motifs regulate the induction of immune genes in Drosophila. J. Mol. Biol. 232:327–33 Lemaitre B, Nicolas E, Michaut L, Reichhart JM, Hoffmann JA. 1996. The dorsoventral regulatory gene cassette spatzle/ Toll/cactus controls the potent antifungal response in Drosophila adults. Cell 86:973–83 Lemaitre B, Reichhart JM, Hoffmann JA. 1997. Drosophila host defense: differential induction of antimicrobial peptide genes after infection by various classes of microorganisms. Proc. Natl. Acad. Sci. USA 94:14614–19 Ip YT, Reach M, Engstrom Y, Kadalayil L, Cai H, Gonzalez-Crespo S, Tatei K, Levine M. 1993. Dif, a dorsal-related gene that mediates an immune response in Drosophila. Cell 75:753–63 Meng X, Khanuja BS, Ip YT. 1999. Toll receptor-mediated Drosophila immune response requires Dif, an NFkappaB factor. Genes Dev. 13:792–97 Rutschmann S, Jung AC, Hetru C, Reichhart JM, Hoffmann JA, Ferrandon D. 2000. The Rel protein DIF mediates the antifungal but not the antibacterial host

1 Feb 2002

10:13

AR

AR152-08.tex

AR152-08.SGM

LaTeX2e(2001/05/10)

P1: GJC

INNATE IMMUNE RECOGNITION

43.

Annu. Rev. Immunol. 2002.20:197-216. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

44.

45.

46.

47.

48.

49.

50.

51.

defense in Drosophila. Immunity 12:569– 80 Levashina EA, Langley E, Green C, Gubb D, Ashburner M, Hoffmann JA, Reichhart JM. 1999. Constitutive activation of toll-mediated antifungal defense in serpin-deficient Drosophila. Science 285:1917–19 Imler JL, Hoffmann JA. 2000. Signaling mechanisms in the antimicrobial host defense of Drosophila. Curr. Opin. Microbiol. 3:16–22 Lemaitre B, Kromer-Metzger E, Michaut L, Nicolas E, Meister M, Georgel P, Reichhart JM, Hoffmann JA. 1995. A recessive mutation, immune deficiency (Imd), defines two distinct control pathways in the Drosophila host defense. Proc. Natl. Acad. Sci. USA 92:9465–69 Hedengren M, Asling B, Dushay MS, Ando I, Ekengren S, Wihlborg M, Hultmark D. 1999. Relish, a central factor in the control of humoral but not cellular immunity in Drosophila. Mol. Cell. 4:827– 37 Lu Y, Wu LP, Anderson KV. 2001. The antibacterial arm of the drosophila innate immune response requires an IkappaB kinase. Genes Dev. 15:104–10 Silverman N, Zhou R, Stoven S, Pandey N, Hultmark D, Maniatis T. 2000. A drosophila IkappaB kinase complex required for relish cleavage and antibacterial immunity. Genes Dev. 14:2461–71 Elrod-Erickson M, Mishra S, Schneider D. 2000. Interactions between the cellular and humoral immune responses in Drosophila. Curr. Biol. 10:781–84 Leulier F, Rodriguez A, Khush RS, Abrams JM, Lemaitre B. 2000. The Drosophila caspase Dredd is required to resist gram-negative bacterial infection. EMBO Rep. 1:353–58 Rutschmann S, Jung AC, Zhou R, Silverman N, Hoffmann JA, Ferrandon D. 2000. Role of Drosophila IKK gamma in a toll-independent antibacterial immune response. Nat. Immunol. 1:342–47

213

52. Chen P, Rodriguez A, Erskine R, Thach T, Abrams JM. 1998. Dredd, a novel effector of the apoptosis activators reaper, grim, and hid in Drosophila. Dev. Biol. 201:202–16 53. Yamaoka S, Courtois G, Bessia C, Whiteside ST, Weil R, Agou F, Kirk HE, Kay RJ, Israel A. 1998. Complementation cloning of NEMO, a component of the IkappaB kinase complex essential for NF-kappaB activation. Cell 93:1231–40 54. Dushay MS, Asling B, Hultmark D. 1996. Origins of immunity: Relish, a compound Rel-like gene in the antibacterial defense of Drosophila. Proc. Natl. Acad. Sci. USA 93:10343–47 55. Tauszig S, Jouanguy E, Hoffmann JA, Imler JL. 2000. From the cover: tollrelated receptors and the control of antimicrobial peptide expression in Drosophila. Proc. Natl. Acad. Sci. USA 97:10520–25 56. Williams MJ, Rodriguez A, Kimbrell DA, Eldon ED. 1997. The 18-wheeler mutation reveals complex antibacterial gene regulation in Drosophila host defense. EMBO J. 16:6120–30 57. Medzhitov R, Preston-Hurlburt P, Janeway CA Jr. 1997. A human homologue of the Drosophila Toll protein signals activation of adaptive immunity. Nature 388:394–97 58. Poltorak A, He X, Smirnova L, Liu MY, Huffel CV, Du X, Birdwell D, Alejos E, Silva M, Galanos C, Freudenberg M, Ricciardi-Castagnoli P, Layton B, Beutler B. 1998. Defective LPS signaling in C3H/HeJ and C57BL/10ScCr mice: mutations in Tlr4 gene. Science 282:2085–88 59. Qureshi ST, Lariviere L, Leveque G, Clermont S, Moore KJ, Gros P, Malo D. 1999. Endotoxin-tolerant mice have mutations in Toll-like receptor 4 (Tlr4) [published erratum appears in J. Exp. Med. 1999, May 3;189(9):following 1518]. J. Exp. Med. 189:615–25 60. Hoshino K, Takeuchi O, Kawai T, Sanjo H, Ogawa T, Takeda Y, Takeda K, Akira S. 1999. Cutting edge: Toll-like receptor

1 Feb 2002

10:13

214

61.

Annu. Rev. Immunol. 2002.20:197-216. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

62.

63.

64.

65.

66.

67.

AR

AR152-08.tex

JANEWAY

¥

AR152-08.SGM

LaTeX2e(2001/05/10)

P1: GJC

MEDZHITOV

4 (TLR4)-deficient mice are hyporesponsive to lipopolysaccharide: evidence for TLR4 as the Lps gene product. J. Immunol. 162:3749–52 Ulevitch RJ, Tobias PS. 1995. Receptordependent mechanisms of cell stimulation by bacterial endotoxin. Annu. Rev. Immunol. 13:437–57 Moore KJ, Andersson LP, Ingalls RR, Monks BG, Li R, Arnaout MA, Golenbock DT, Freeman MW. 2000. Divergent response to LPS and bacteria in CD14deficient murine macrophages. J. Immunol. 165:4272–80 Haziot A, Ferrero E, Kontgen F, Hijiya N, Yamamoto S, Silver J, Stewart CL, Goyert SM. 1996. Resistance to endotoxin shock and reduced dissemination of gramnegative bacteria in CD14-deficient mice. Immunity 4:407–14 Shimazu R, Akashi S, Ogata H, Nagai Y, Fukudome K, Miyake K, Kimoto M. 1999. MD-2, a molecule that confers lipopolysaccharide responsiveness on Toll-like receptor 4. J. Exp. Med. 189: 1777–82 Schromm AB, Lien E, Henneke P, Chow JC, Yoshimura A, Heine H, Latz E, Monks BG, Schwartz DA, Miyake K, Golenbock DT. 2001. Molecular genetic analysis of an endotoxin nonresponder mutant cell line. A point mutation in a conserved region of md-2 abolishes endotoxin-induced signaling. J. Exp. Med. 194:79–88 Lien E, Means TK, Heine H, Yoshimura A, Kusumoto S, Fukase K, Fenton MJ, Oikawa M, Qureshi N, Monks B, Finberg RW, Ingalls RR, Golenbock DT. 2000. Toll-like receptor 4 imparts ligand-specific recognition of bacterial lipopolysaccharide. J. Clin. Invest. 105:497–504 da Silva Correia J, Soldau K, Christen U, Tobias PS, Ulevitch RJ. 2001. Lipopolysaccharide is in close proximity to each of the proteins in its membrane receptor complex. Transfer from cd14 to tlr4 and md-2. J. Biol. Chem. 276:21129–35

68. Poltorak A, Ricciardi-Castagnoli P, Citterio S, Beutler B. 2000. Physical contact between lipopolysaccharide and toll-like receptor 4 revealed by genetic complementation. Proc. Natl. Acad. Sci. USA 97: 2163–67 69. Miyake K, Yamashita Y, Ogata M, Sudo T, Kimoto M. 1995. RP105, a novel B cell surface molecule implicated in B cell activation, is a member of the leucinerich repeat protein family. J. Immunol. 154:3333–40 70. Miura Y, Shimazu R, Miyake K, Akashi S, Ogata H, Yamashita Y, Narisawa Y, Kimoto M. 1998. RP105 is associated with MD-1 and transmits an activation signal in human B cells. Blood 92:2815– 22 71. Miyake K, Shimazu R, Kondo J, Niki T, Akashi S, Ogata H, Yamashita Y, Miura Y, Kimoto M. 1998. Mouse MD-1, a molecule that is physically associated with RP105 and positively regulates its expression. J. Immunol. 161:1348–53 72. Ogata H, Su I, Miyake K, Nagai Y, Akashi S, Mecklenbrauker I, Rajewsky K, Kimoto M, Tarakhovsky A. 2000. The toll-like receptor protein RP105 regulates lipopolysaccharide signaling in B cells. J. Exp. Med. 192:23–29 73. Takeuchi O, Hoshino K, Kawai T, Sanjo H, Takada H, Ogawa T, Takeda K, Akira S. 1999. Differential roles of TLR2 and TLR4 in recognition of gram-negative and gram-positive bacterial cel wall components. Immunity 11:443–51 74. Vabulas RM, Ahmad-Nejad P, da Costa C, Miethke T, Kirschning CJ, Hacker H, Wagner H. 2001. Endocytosed heat shock protein 60s use TLR2 and TLR4 to activate the toll/interleukin-1 receptor signaling pathway in innate immune cells. J. Biol. Chem. 11:11 75. Ohashi K, Burkart V, Flohe S, Kolb H. 2000. Cutting edge: Heat shock protein 60 is a putative endogenous ligand of the toll-like receptor-4 complex. J. Immunol. 164:558–61

1 Feb 2002

10:13

AR

AR152-08.tex

AR152-08.SGM

LaTeX2e(2001/05/10)

P1: GJC

Annu. Rev. Immunol. 2002.20:197-216. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

INNATE IMMUNE RECOGNITION 76. Kurt-Jones EA, Popova L, Kwinn L, Haynes LM, Jones LP, Tripp RA, Walsh EE, Freeman MW, Golenbock DT, Anderson LJ, Finberg RW. 2000. Pattern recognition receptors TLR4 and CD14 mediate response to respiratory syncytial virus. Nat. Immunol. 1:398–401 77. Schwandner R, Diarski R, Wesche H, Rothe M, Kirschning CJ. 1999. Peptidoglycan- and lipoteichoic acid-induced cell activation is mediated by toll-like receptor 2. J. Biol. Chem. 274:17406– 9 78. Aliprantis AO, Yang RB, Mark MR, Suggett S, Devaux B, Radolf JD, Klimpel GR, Godowski P, Zychlinsky A. 1999. Cell activation and apoptosis by bacterial lipoproteins through toll-like receptor-2. Science 285:736–39 79. Brightbill HD, Libraty DH, Krutzik SR, Yang RB, Belisle JT, Bleharski JR, Maitland M, Norgard MV, Plevy SE, Smale ST, Brennan PJ, Bloom BR, Godowski PJ, Modlin RL. 1999. Host defense mechanisms triggered by microbial lipoproteins through toll-like receptors. Science 285:732–36 80. Takeuchi O, Kaufmann A, Grote K, Kawai T, Hoshino K, Morr M, Muhlradt PF, Akira S. 2000. Cutting edge: Preferentially the R-stereoisomer of the mycoplasmal lipopeptide macrophage-activating lipopeptide-2 activates immune cells through a toll-like receptor 2- and MyD88-dependent signaling pathway. J. Immunol. 164:554–57 81. Hajjar AM, O’Mahony DS, Ozinsky A, Underhill DM, Aderem A, Klebanoff SJ, Wilson CB. 2001. Cutting edge: functional interactions between toll-like receptor (TLR) 2 and TLR1 or TLR6 in response to phenol-soluble modulin. J. Immunol. 166:15–19 82. Hirschfeld M, Weis JJ, Toshchakov V, Salkowski CA, Cody MJ, Ward DDC, Qureshi N, Michalek SM, Vogel SN. 2001. Signaling by toll-like receptor 2 and 4 agonists results in differential gene

83.

84.

85.

86.

87.

88.

89.

90.

215

expression in murine macrophages. Infect. Immun. 69:1477–82 Werts C, Tapping RI, Mathison JC, Chuang TH, Kravchenko V, Saint Girons I, Haake DA, Godowski PJ, Hayashi F, Ozinsky A, Underhill DM, Kirschning CJ, Wagner H, Aderem A, Tobias PS, Ulevitch RJ. 2001. Leptospiral lipopolysaccharide activates cells through a TLR2-dependent mechanism. Nat. Immunol. 2:346–52 Campos MA, Almeida IC, Takeuchi O, Akira S, Valente EP, Procopio DO, Travassos LR, Smith JA, Golenbock DT, Gazzinelli RT. 2001. Activation of Tolllike receptor-2 by glycosylphosphatidylinositol anchors from a protozoan parasite. J. Immunol. 167:416–23 Underhill D, Ozinsky A, Hajjar A, Stevens A, Wilson C, Bassetti M, Aderem A. 1999. The Toll-like receptor 2 is recruited to macrophage phagosomes and discriminates between pathogens. Nature 401:811–15 Ozinsky A, Underhill DM, Fontenot JD, Hajjar AM, Smith KD, Wilson CB, Schroeder L, Aderem A. 2000. The repertoire for pattern recognition of pathogens by the innate immune system is defined by cooperation between toll-like receptors. Proc. Natl. Acad. Sci. USA 97:13766–71 Takeuchi O, Kawai T, Muhlradt PF, Morr M, Radolf JD, Zychlinsky A, Takeda K, Akira S. 2001. Discrimination of bacterial lipoproteins by Toll-like receptor 6. Int. Immunol. 13:933–40 Hayashi F, Smith KD, Ozinsky A, Hawn TR, Yi EC, Goodlett DR, Eng JK, Akira S, Underhill DM, Aderem A. 2001. The innate immune response to bacterial flagellin is mediated by Toll-like receptor 5. Nature 410:1099–103 Krieg AM, Yi AK, Matson S, Waldschmidt TJ, Bishop GA, Teasdale R, Koretzky GA, Klinman DM. 1995. CpG motifs in bacterial DNA trigger direct B-cell activation. Nature 374:546–49 Krieg AM. 2000. The role of CpG motifs

1 Feb 2002

10:13

216

91.

Annu. Rev. Immunol. 2002.20:197-216. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

92.

93.

94.

95.

96.

97.

98.

99.

AR

AR152-08.tex

JANEWAY

¥

AR152-08.SGM

LaTeX2e(2001/05/10)

P1: GJC

MEDZHITOV

in innate immunity. Curr. Opin. Immunol. 12:35–43 Hemmi H, Takeuchi O, Kawai T, Kaisho T, Sato S, Sanjo H, Matsumoto M, Hoshino K, Wagner H, Takeda K, Akira S. 2000. A Toll-like receptor recognizes bacterial DNA. Nature 408:740–45 Hacker H, Mischak H, Miethke T, Liptay S, Schmid R, Sparwasser T, Heeg K, Lipford GB, Wagner H. 1998. CpG-DNAspecific activation of antigen-presenting cells requires stress kinase activity and is preceded by non-specific endocytosis and endosomal maturation. EMBO J. 17:6230–40 Kopp EB, Medzhitov R. 1999. The Tollreceptor family and control of innate immunity. Curr. Opin. Immunol. 11:13–18 Medzhitov R, Preston-Hurlburt P, Kopp E, Stadlen A, Chen C, Ghosh S, Janeway CA Jr. 1998. MyD88 is an adaptor protein in the hToll/IL-1 receptor family signaling pathways. Mol. Cell. 2:253–58 Muzio M, Natoli G, Saccani S, Levrero M, Mantovani A. 1998. The human Toll signaling pathway: divergence of nuclear factor kappaB and JNK/SAPK activation upstream of tumor necrosis factor receptor-associated factor 6 (TRAF6). J. Exp. Med. 187:2097–101 Wesche H, Henzel WJ, Shillinglaw W, Li S, Cao Z. 1997. MyD88: an adapter that recruits IRAK to the IL-1 receptor complex. Immunity 7:837–47 Burns K, Clatworthy J, Martin L, Martinon F, Plumpton C, Maschera B, Lewis A, Ray K, Tschopp J, Volpe F. 2000. Tollip, a new component of the IL-1RI pathway, links IRAK to the IL-1 receptor. Nat. Cell. Biol. 2:346–51 Cao Z, Xiong J, Takeuchi M, Kurama T, Goeddel DV. 1996. RAF6 is a signal transducer for interleukin-1. Nature 383:443– 46 Deng L, Wang C, Spencer E, Yang L, Braun A, You J, Slaughter C, Pickart C, Chen ZJ. 2000. Activation of the Ikap-

100.

101.

102.

103.

104.

105.

106.

107.

108.

paB kinase complex by TRAF6 requires a dimeric ubiquitin-conjugating enzyme complex and a unique polyubiquitin chain. Cell 103:351–61 Arbibe L, Mira JP, Teusch N, Kline L, Guha M, Mackman N, Godowski PJ, Ulevitch RJ, Knaus UG. 2000. Toll-like receptor 2-mediated NF-kappa B activation requires a Rac1-dependent pathway. Nat. Immunol. 1:533–40 Kawai T, Adachi O, Ogawa T, Takeda K, Akira S. 1999. Unresponsiveness of MyD88-deficient mice to endotoxin. Immunity 11:115–22 Kaisho T, Takeuchi O, Kawai T, Hoshino K, Akira S. 2001. Endotoxin-induced maturation of myd88-deficient dendritic cells. J. Immunol. 166:5688–94 Adachi O, Kawai T, Takeda K, Matsumoto M, Tsutsui H, Sakagami M, Nakanishi K, Akira S. 1998. Targeted disruption of the MyD88 gene results in loss of IL-1- and IL-18-mediated function. Immunity 9:143–50 Schnare M, Holt AC, Takeda K, Akira S, Medzhitov R. 2000. Recognition of CpG DNA is mediated by signaling pathways dependent on the adaptor protein MyD88. Curr. Biol. 10:1139–42 Horng T, Barton G, Medzhitov R. 2001. TIRAP, a novel adapter in the Toll signaling pathway. Nat. Immunol. 2:1139– 42 Schnare M, Barton G, Takeda K, Akira S, Medzhitov R. 2001. Role of Toll-like receptors in the control of adaptive immune responses. Nat. Immunol. 2:In press Means TK, Wang S, Lien E, Yoshimura A, Golenbock DT, Fenton MJ. 1999. Human toll-like receptors mediate cellular activation by Mycobacterium tuberculosis. J. Immunol. 163:3920–27 Underhill DM, Ozinsky A, Smith KD, Aderem A. 1999. Toll-like receptor-2 mediates mycobacteria-induced proinflammatory signaling in macrophages. Proc. Natl. Acad. Sci. USA 96:14459–63

P1: FRK

February 1, 2002

12:47

Annual Reviews

AR152-FM

Annual Review of Immunology Volume 20, 2002

Annu. Rev. Immunol. 2002.20:197-216. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

CONTENTS FRONTISPIECE, Charles A. Janeway, Jr. A TRIP THROUGH MY LIFE WITH AN IMMUNOLOGICAL THEME, Charles A. Janeway, Jr.

xiv 1

THE B7 FAMILY OF LIGANDS AND ITS RECEPTORS: NEW PATHWAYS FOR COSTIMULATION AND INHIBITION OF IMMUNE RESPONSES, Beatriz M. Carreno and Mary Collins

29

MAP KINASES IN THE IMMUNE RESPONSE, Chen Dong, Roger J. Davis, and Richard A. Flavell

PROSPECTS FOR VACCINE PROTECTION AGAINST HIV-1 INFECTION AND AIDS, Norman L. Letvin, Dan H. Barouch, and David C. Montefiori T CELL RESPONSE IN EXPERIMENTAL AUTOIMMUNE ENCEPHALOMYELITIS (EAE): ROLE OF SELF AND CROSS-REACTIVE ANTIGENS IN SHAPING, TUNING, AND REGULATING THE AUTOPATHOGENIC T CELL REPERTOIRE, Vijay K. Kuchroo, Ana C. Anderson, Hanspeter Waldner, Markus Munder, Estelle Bettelli, and Lindsay B. Nicholson

55 73

101

NEUROENDOCRINE REGULATION OF IMMUNITY, Jeanette I. Webster, Leonardo Tonelli, and Esther M. Sternberg

125

MOLECULAR MECHANISM OF CLASS SWITCH RECOMBINATION: LINKAGE WITH SOMATIC HYPERMUTATION, Tasuku Honjo, Kazuo Kinoshita, and Masamichi Muramatsu

165

INNATE IMMUNE RECOGNITION, Charles A. Janeway, Jr. and Ruslan Medzhitov

KIR: DIVERSE, RAPIDLY EVOLVING RECEPTORS OF INNATE AND ADAPTIVE IMMUNITY, Carlos Vilches and Peter Parham ORIGINS AND FUNCTIONS OF B-1 CELLS WITH NOTES ON THE ROLE OF CD5, Robert Berland and Henry H. Wortis E PROTEIN FUNCTION IN LYMPHOCYTE DEVELOPMENT, Melanie W. Quong, William J. Romanow, and Cornelis Murre

197 217 253 301

LYMPHOCYTE-MEDIATED CYTOTOXICITY, John H. Russell and Timothy J. Ley

SIGNAL TRANSDUCTION MEDIATED BY THE T CELL ANTIGEN x

323

P1: FRK

Annu. Rev. Immunol. 2002.20:197-216. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

February 1, 2002

12:47

Annual Reviews

AR152-FM

CONTENTS

xi

RECEPTOR: THE ROLE OF ADAPTER PROTEINS, Lawrence E. Samelson INTERACTION OF HEAT SHOCK PROTEINS WITH PEPTIDES AND ANTIGEN PRESENTING CELLS: CHAPERONING OF THE INNATE AND ADAPTIVE IMMUNE RESPONSES, Pramod Srivastava CHROMATIN STRUCTURE AND GENE REGULATION IN THE IMMUNE SYSTEM, Stephen T. Smale and Amanda G. Fisher PRODUCING NATURE’S GENE-CHIPS: THE GENERATION OF PEPTIDES FOR DISPLAY BY MHC CLASS I MOLECULES, Nilabh Shastri, Susan

371

Schwab, and Thomas Serwold

395 427

463

THE IMMUNOLOGY OF MUCOSAL MODELS OF INFLAMMATION, Warren Strober, Ivan J. Fuss, and Richard S. Blumberg T CELL MEMORY, Jonathan Sprent and Charles D. Surh

GENETIC DISSECTION OF IMMUNITY TO MYCOBACTERIA: THE HUMAN MODEL, Jean-Laurent Casanova and Laurent Abel ANTIGEN PRESENTATION AND T CELL STIMULATION BY DENDRITIC CELLS, Pierre Guermonprez, Jenny Valladeau, Laurence Zitvogel, Clotilde Th´ery, and Sebastian Amigorena

495 551 581

621

NEGATIVE REGULATION OF IMMUNORECEPTOR SIGNALING, Andr´e Veillette, Sylvain Latour, and Dominique Davidson

669

CPG MOTIFS IN BACTERIAL DNA AND THEIR IMMUNE EFFECTS, Arthur M. Krieg

PROTEIN KINASE Cθ

709 IN

T CELL ACTIVATION, Noah Isakov and Amnon

Altman

RANK-L AND RANK: T CELLS, BONE LOSS, AND MAMMALIAN EVOLUTION, Lars E. Theill, William J. Boyle, and Josef M. Penninger PHAGOCYTOSIS OF MICROBES: COMPLEXITY IN ACTION, David M. Underhill and Adrian Ozinsky

761 795 825

STRUCTURE AND FUNCTION OF NATURAL KILLER CELL RECEPTORS: MULTIPLE MOLECULAR SOLUTIONS TO SELF, NONSELF DISCRIMINATION, Kannan Natarajan, Nazzareno Dimasi, Jian Wang, Roy A. Mariuzza, and David H. Margulies

853

INDEXES Subject Index Cumulative Index of Contributing Authors, Volumes 1–20 Cumulative Index of Chapter Titles, Volumes 1–20

ERRATA An online log of corrections to Annual Review of Immunology chapters (if any, 1997 to the present) may be found at http://immunol.annualreviews.org/

887 915 925

8 Feb 2002

9:43

AR

AR152-09.tex

AR152-09.SGM

LaTeX2e(2001/05/10) P1: GJC 10.1146/annurev.immunol.20.092501.134942

Annu. Rev. Immunol. 2002. 20:217–51 DOI: 10.1146/annurev.immunol.20.092501.134942 c 2002 by Annual Reviews. All rights reserved Copyright °

KIR: Diverse, Rapidly Evolving Receptors of Innate and Adaptive Immunity Carlos Vilches1 and Peter Parham2 Annu. Rev. Immunol. 2002.20:217-251. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

1

Servicio de Inmunolog´ıa, Hospital Universitario Cl´ınica Puerta de Hierro, San Mart´ın de Porres 4, 28035 Madrid, Spain; e-mail: [email protected] 1 Departments of Structural Biology and Microbiology & Immunology, Stanford University, Stanford, California 94305-5126; e-mail: [email protected]

Key Words evolution, gene diversity, MHC, NK cells, polymorphism ■ Abstract KIR genes have evolved in primates to generate a diverse family of receptors with unique structures that enable them to recognize MHC-class I molecules with locus and allele-specificity. Their combinatorial expression creates a repertoire of NK cells that surveys the expression of almost every MHC molecule independently, thus antagonizing the spread of pathogens and tumors that subvert innate and adaptive defense by selectively downregulating certain MHC class I molecules. The genes encoding KIR that recognize classical MHC molecules have diversified rapidly in human and primates; this contrasts with conservation of immunoglobulin- and lectinlike receptors for nonclassical MHC molecules. As a result of the variable KIR-gene content in the genome and the polymorphism of the HLA system, dissimilar numbers and qualities of KIR:HLA pairs function in different humans. This diversity likely contributes variability to the function of NK cells and T-lymphocytes by modulating innate and adaptive immune responses to specific challenges.

INTRODUCTION MHC Class I Receptors of NK Cells Natural killer (NK) cells participate in early, innate defense through cytotoxic activity against pathogen-infected cells and secretion of cytokines and chemokines that modulate subsequent steps in the adaptive immune response (1, 2). Although NK cells can be activated by dangerous cells in several ways, many details of the mechanisms remain poorly understood (3–10). Best characterized of the events that activate NK cells is reduced cell-surface expression of MHC class I molecules (11, 12). This trigger is a common consequence of intracellular infection because many pathogens have evolved mechanisms that sabotage MHC class I expression and antigen presentation to CD8 T cells. Such mechanisms allow infected cells to avoid detection by T-lymphocytes (13–15); similarly, the action of tumor-specific 0732-0582/02/0407-0217$14.00

217

8 Feb 2002

8:43

Annu. Rev. Immunol. 2002.20:217-251. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

218

AR

AR152-09.tex

VILCHES

¥

AR152-09.SGM

LaTeX2e(2001/05/10)

P1: GJC

PARHAM

CD8 T cells selects for resistant, variant tumor cells bearing mutations that reduce expression of MHC class I (16). NK cells detect downregulation of MHC class I molecules by means of specific membrane receptors, which under most circumstances prevent NK cells from responding to healthy cells expressing the normal complement of autologous MHC class I molecules. Two main categories of these inhibitory receptors have been defined in humans—the CD94:NKG2A heterodimer (17–19), of which both components belong to the C-type lectin-like superfamily, and the killer-cell immunoglobulin-like receptor (KIR) family (20, 21). In addition, some members of the ILT/LIR/MIR family, related evolutionarily to KIR, are also expressed by NK cells and recognize HLA class I molecules (22, 23), but their importance for NK-cell function has yet to be defined. The diversity of MHC class I molecules complicates the job of receptors that survey their levels of expression. The CD94:NKG2A and KIR molecules present contrasting evolutionary strategies of coping with this challenge. The CD94:NKG2A approach has been to ignore MHC class I diversity by recognizing a sequence element conserved in the signal peptide of most HLA class I molecules. Peptides containing this element are then bound by HLA-E, a class I molecule with little polymorphism that cannot present its own leader peptide, and these complexes are the ligands for the nonpolymorphic CD94:NKG2A receptor (17–19, 24). On the other hand, the KIR approach has been to embrace the diversity of MHC class I through direct recognition of polymorphic determinants, by evolving a highly variable and polymorphic KIR system with diversity comparable to that of MHC class I (25). The different approaches taken by CD94:NKG2 and KIR toward a similar purpose is consistent with their being complementary rather than redundant systems. Mouse NK cells also have two categories of inhibitory MHC class I receptors, one of which is a CD94:NKG2A receptor with functions similar to the human receptor (26). Mice do not have KIR; instead they have a diverse family of Ly49 molecules, each having different specificity for MHC class I polymorphisms (27– 29). Ly49 receptors, however, are unrelated phylogenetically to KIR because they belong to the C-type lectin-like superfamily. The mouse-human comparison shows that the CD94:NKG2 category of MHC class I–specific receptor is older and more conserved than either KIR or Ly49. Evolution of KIR and Ly49 has converged on becoming inhibitory NK-cell receptors for MHC class I; this evolution argues that a family of such receptors, each specific for diverse ligands, has advantages and can accomplish functions that receptors focused on monomorphic MHC class I motifs cannot. The limitations of the surveillance mediated by the CD94:NKG2A receptor are perhaps best illustrated by how pathogens subvert it. For instance, the leader peptide of the human cytomegalovirus (HCMV) protein UL40 stabilizes expression of HLA-E, allowing HCMV-infected cells to inhibit any NK cell expressing CD94:NKG2A (30). Even more effective is a strategy that combines downregulation of MHC class I molecules that present immunodominant viral epitopes

8 Feb 2002

8:43

AR

AR152-09.tex

AR152-09.SGM

LaTeX2e(2001/05/10)

P1: GJC

Annu. Rev. Immunol. 2002.20:217-251. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

KIR DIVERSITY

219

to cytolytic T cells (CTL) with sustained expression of MHC class I ligands that inhibit NK cells. The human immunodeficiency virus (HIV) does this using the Nef protein, which downregulates HLA-A and -B in infected cells while preserving expression of HLA-C and E (31, 32), a combination that provides inhibitory ligands for most if not all NK cells. Similarly, the HCMV proteins US2 and US11 target HLA-A and -B for proteosome degradation, but not HLA-E or -G (their effect on HLA-C being controversial) (33–35). Herpes simplex virus 1 (HSV-1) has also been proposed to exert differential effects on the expression of HLA-A, -B, and -C molecules (36). In principle, a family of receptors, each specific for a different subgroup of MHC molecules and expressed by only a subset of NK cells, could enable the expression of each MHC molecule to be surveyed, thus antagonizing the spread of infection by pathogens whose evasion is based upon selective downregulation of particular MHC class I molecules. In aspiring but hardly complete fashion, these are the features of the primate KIR and rodent Ly49 systems of receptors.

STRUCTURAL DIVERSITY OF KIR: GENETIC BASIS Today the human KIR family is represented in gene data banks by more than one hundred mRNA and DNA sequences. Three criteria—number of extracellular Ig-like domains, cytoplasmic tail length, and sequence similarity—have been used to classify the encoded KIR proteins into 13 groups. In the nomenclature used to describe these groups, KIR3DL1–2, KIR3DS1, KIR2DL1–5, and KIR2DS1–5, the number of Ig-like domains is given by 2D for 2 domains or 3D for 3 domains; the length of the cytoplasmic tails is given as L for long or S for short; and different KIR with similar overall organization but sequence divergence of >2% are generally numbered in series (Table 1). The inhibitory KIR have long cytoplasmic tails containing pairs of immune tyrosine-based inhibitory motifs (ITIMs), whereas the KIR with short cytoplasmic tails are activating receptors that associate with the DAP12 signaling molecule via a positively charged lysine residue in their transmembrane domain (37). The prototypical KIR from which all others can be derived is a long-tailed KIR with three extracellular Ig-like domains, represented in humans by KIR3DL1 and KIR3DL2. These KIR are encoded by separate genes that span 14–16 kb and are organized in nine exons, which roughly correspond to different functional regions of the protein (38) (Figure 1). Exons 1 and 2 code for the signal peptide plus the first two amino acids of the mature polypeptide. Exons 3, 4, and 5 each encode an Ig-like domain, named D0, D1, and D2, respectively. A stem, encoded by exon 6, connects the D2 domain to the transmembrane region encoded by exon 7. The cytoplasmic tail is encoded by exons 8 and 9. The majority of human KIR have two extracellular Ig-like domains. These are of two types: type 1 KIR2D having domains homologous to D1 and D2 of KIR3DL, and type 2 KIR2D having domains homologous to D0 and D2 of KIR3D. Genes

8 Feb 2002

8:43

AR

220

AR152-09.tex

VILCHES

¥

AR152-09.SGM

LaTeX2e(2001/05/10)

P1: GJC

PARHAM

Annu. Rev. Immunol. 2002.20:217-251. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

TABLE 1 Basic structural and functional features of KIR

KIR

Extracellular region (Ig-like domains)

Charged amino acid in tm region

Cytoplasmic tail (a.a.); ITIMs

Known ligands

mRNA

3DL1

D0-D1-D2

—

84; 2

HLA-BBw4

+

3DL2

D0-D1-D2

—

95; 2

HLA-A3, others?

+

3DS1

D0-D1-D2

Lys

22,27; 0

?

+ Lys80

+

2DL1

D1-D2

—

84; 2

HLA-C

2DL2,3

D1-D2

—

84,76; 2

HLA-CAsn80

+

Lys80

2DS1

D1-D2

Lys

39; 0

HLA-C

+

2DS2

D1-D2

Lys

39; 0

HLA-CAsn80

+

2DS3,5

D1-D2

Lys

39; 0

?

+

2DS4

D1-D2

Lys

39; 0

HLA-C?

+

2DL4

D0-D2

Arg

115; 1 or 11; 0

HLA-G

+

2DL5

D0-D2

—

115; 2

?

+(−)a

KIRC1

D0-D1-D2 (no stem)

—

67; 1

?

−(+)b

a

Some KIR2DL5 variants are not transcribed.

b

Although generally nontranscribed, a cDNA sequence for KIRC1 has been recently deposited in the GenBank (47a).

encoding type 1 KIR2D are similar in organization to those encoding KIR3D and contain a region homologous to exon 3 encoding the D0 domain of KIR3D (78.7– 79.8% sequence identity with 3DL1 and 79.8–81.2% with 3DL2) (39, 40). This region, called “pseudoexon 3,” is spliced out of the RNA transcript even when it maintains the correct reading frame and has correct splicing sites, as is the case in half of the genes encoding type 1 KIR2D (39). The one feature that distinguishes exons 3 from the pseudoexons 3 is that they are longer by one codon. If the three nucleotides missing from pseudoexons are part of an exonic splicing enhancer necessary for inclusion of exon 3 in mature mRNA (39), then their absence could

Figure 1 Organization of the KIR3DL1 gene (38). Exons (boxes) and introns (lines) are represented approximately to scale.

8 Feb 2002

8:43

AR

AR152-09.tex

AR152-09.SGM

LaTeX2e(2001/05/10)

P1: GJC

Annu. Rev. Immunol. 2002.20:217-251. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

KIR DIVERSITY

221

explain the lack of expression of type 1 KIR2D pseudoexons. In humans, type 2 KIR2D are represented by 2DL4 and 2DL5. In comparison to genes encoding KIR3D, the 2DL4 and 2DL5 genes have a deletion >2 kb that includes exon 4 encoding the D1 domain (41, 42). KIR cytoplasmic tails can have different lengths, despite that exons encoding them have similar lengths and sequences. This is because single-nucleotide substitutions or short indels produce stop codons at diverse positions within the exonic sequence (43). The varied nature of the polymorphisms leading to short cytoplasmic tails is one of several striking examples of convergent evolution encountered in studying NK-cell receptors for MHC class I. It argues strongly in favor of there being important functions for activating KIR. No other major variations in the reading frames of functional KIR genes have been described. All other polymorphic structural features, with known or suspected importance for MHC class I specificity or signal transduction, are due to simple nucleotide replacements. Among these are ■

■

■

Charged residues in the transmembrane region: lysine in activating KIR and arginine at a different position in KIR2DL4. This and the following feature suggest that 2DL4 may be an activating receptor despite its long cytoplasmic tail (44). Indeed, ligation of KIR2DL4 has been shown to activate production of interferon-γ , but not cytotoxicity (44a). Inactivation of the carboxy-terminal ITIM in 2DL4 due to substitution of cysteine for tyrosine. Residues conferring specificity for HLA-C allotypes to 2DL1–3 and activating KIR2D.

In addition to the ∼13 expressed genes, three other KIR genes or gene fragments (usually not transcribed) have been defined, and for these a uniform nomenclature has yet to be adopted. The structure of KIRC1 (45), also known as KIR44 (46) or 3DL3 (47), is similar to that of KIR3D genes, except for absence of a stem-encoding exon and presence of a stop codon before the second ITIM. It is expressed at a very low level (47a). The nucleotide sequence of KIRC1 is phylogenetically about equidistant from KIR3D and the two types of KIR2D (sequence identity to other KIR in exons 1–9: 88.2% to 3DL1, 89.7% to 2DL5.1, 86.3 to 2DL4, 91.1% to 2DL3). Further, phylogenetic analysis does not consistently assign KIRC1 to any of the aforementioned KIR groups (38, 46, 48) (Figure 2); rather, it suggests that this gene defines a different KIR lineage. KIR48 (46), also called KIRX (47) or 2DS6 (38), is a gene fragment lacking exons 6–9 (also lacking exon 2 in many haplotypes). The KIR48 sequence is close to those of type 1 KIR2D, but its third exon lacks the three-nucleotide deletion characteristic of pseudoexons in the latter genes. KIR48 could thus retain features of the ancestral KIR3D gene from which type 1 KIR2D genes were derived (39). Finally, KIR15 is an inactivated type 1 KIR2D gene (46) and could correspond to the gene fragment known as KIRZ (47) or KIRY (49). Other genes and cDNA sequences corresponding to KIR15 indicate

8 Feb 2002

8:43

Annu. Rev. Immunol. 2002.20:217-251. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

222

AR

AR152-09.tex

VILCHES

¥

AR152-09.SGM

LaTeX2e(2001/05/10)

P1: GJC

PARHAM

Figure 2 Phylogenetic relationships among representative KIR genes of four primate species. A distance tree was constructed by a neighbor-joining method (77) using the Pileup and PAUP search applications of the Wisconsin package (Genetics Computer Group), and this tree is represented using TreeView (by Roderic D.M. Page). Confidence of groupings was estimated by 1000 bootstrap replicates (78), and its percentage is shown besides each branch. The following sequences were analyzed: Homo sapiens 2DL3 (L76662), 2DL4 (AF034773), 2DL5 (AF20903), 3DL1 (L41269), KIRC1 cDNA (AF352324), KIRX (exons 2–5 from AC011501); Pan troglodytes 2DL4 (AF258804), 2DL5 (AF258805), 2DL6 (AF258806), 3DL1/2 (AF258798), 3DL4 (AF258800); Pan paniscus 2DL4 (AF266736), 3DLa (AF266732), 3DL4 (AF266731); Macaca mulatta 2DL4 (AF334644), 2DL5 (AF334646), 3DL1 (AF334616), 3DL6 (AF334621). For the sake of sequence-length homogeneity, sequences 50 and 30 of the regions homologous to the motifs AGGGCCGGTC (exon 2) and TCTAGGGAGA (exon 9) of Hs-3DL1, respectively, were trimmed in all entries. KIRC1 was excluded from KIR-lineage 1 due to its different exon organization and variable phylogenetic clustering with different other KIR genes (38, 46, 48). Similarly, KIR3D of the rhesus monkey have not been included in lineage 2 due to the low confidence of its grouping with genes of this lineage.

8 Feb 2002

8:43

AR

AR152-09.tex

AR152-09.SGM

LaTeX2e(2001/05/10)

P1: GJC

KIR DIVERSITY

223

that some alleles of this locus are not fragments but complete genes (38). What all the sequences share is a frame-shift in exon 4.

VARIABLE ORGANIZATION OF THE KIR-GENE COMPLEX

Annu. Rev. Immunol. 2002.20:217-251. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

KIR-Haplotype Diversity In humans, KIR are encoded by a compact family of genes that occupies ∼150 kb of the leukocyte receptor cluster (LRC) on chromosome 19q13.4 and contains no other types of gene (38, 47, 49, 50). A characteristic feature of KIR haplotypes is variability in the quantity and quality of the genes they contain (47, 51). Of the nine and ten genes and pseudogenes present in two sequenced KIR haplotypes, only two expressed genes (2DL4 and 3DL2), one unexpressed gene (KIRC1), and one gene fragment (KIRX) are held in common (47). Genomic typing in populations and in families is consistent with 2DL4, 3DL2, KIRC1, and KIRX being the common components of KIR haplotypes (46, 51–54). Inferred from the diverse KIR genotypes obtained in such studies are many additional KIR haplotypes from the two sequenced, which differ in the presence, absence, duplication, and hybridization of particular genes. The variable number of genes in KIR haplotypes is likely the result of extensive gene duplications and nonreciprocal crossing-over events that are facilitated by the sequence similarity of KIR introns and intergenic regions and the proximity of the genes (38, 47, 50). Of the genes common to KIR haplotypes, KIRC1 and 3DL2 define the ends of the KIR-gene region and KIRX-2DL4, the middle. These genes have been called the framework genes, which then define two intervals containing genes that vary between haplotypes (47, 49). Two distinct forms of haplotype can be distinguished on the basis of several features (Table 2). The A haplotype has fewer genes, with 2DS4 being the only gene encoding a short-tailed KIR (51). B haplotypes have more genes than A haplotypes, and these include 2DL5 as well as various combinations of the genes encoding short-tailed KIR (46, 51, 55). The presence of the 2DL5 gene in B haplotypes is associated with a large (∼24 kb) HindIII fragment on Southern blots that is not produced from A haplotypes. The A haplotypes are characterized by having the 2DL3-KIRZ-2DL1 linkage group in interval one, whereas B haplotypes

TABLE 2 Summary of dimorphisms of typical KIR haplotypes, arranged from centromer to telomer according to Martin et al. (38) and Wilson et al. (40) KIR-cluster interval

A haplotype

“Complete” B haplotype

1

2DL3-KIRZ-2DL1 —

2DL2 2DL5.2

2

3DL1 — 2DS4

3DS1 2DL5.1 Combinations of 2DS1–3, 2DS5

8 Feb 2002

8:43

Annu. Rev. Immunol. 2002.20:217-251. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

224

AR

AR152-09.tex

VILCHES

¥

AR152-09.SGM

LaTeX2e(2001/05/10)

P1: GJC

PARHAM

have 2DL2 and, often, 2DL5.2 (42, 47, 51). In interval two, A haplotypes have 3DL1 and 2DS4, whereas B haplotypes are typified by 3DS1, 2DL5.1, and various combinations of 2DS1–3 and 2DS5 (42, 47, 51, 55). In terms of gene content, the A haplotypes are less variable than B haplotypes, although the frequencies of A and B haplotypes in populations are relatively even (51). Because of the propensity for recombination within the KIR genes, it is likely that certain haplotypes will not place comfortably in either A or B category because they represent hybrids. Also possible is that the gene order and placement in some haplotypes may significantly differ from those published so far (38, 40). From comparison of KIR haplotypes, various consequences of unequal recombination have been appreciated. First is the duplication of genes, as illustrated by the presence of haplotypes having one or two copies of 2DL5 differing by eight or fewer nucleotide substitutions (42, 56; M. G´omez-Lozano, C. Vilches, unpublished). Second is the recombination of one copy of a duplicated gene with other KIR genes to evolve pairs of KIR with similar ligand-binding specificity, but distinct pathways of signal transduction. Examples of such pairs are 2DL1/2DS1 and 2DL2/2DS2. When a similar recombination involves a single-copy gene, it can create alleles of the locus that differ in their potential for signal transduction. That appears to be the case for 3DL1 and 3DS1, which generally segregate as alleles (47, 51, 53, 54). The mutual exclusivity of the genes 2DL3-KIRZ-2DL1 (A haplotypes) and 2DL2 (B haplotypes) can also be attributed to an operational allelism caused by asymmetric recombination. 2DL2 is a hybrid produced by recombining the centromeric part of 2DL3 with the telomeric part of 2DL1, an event that would have deleted the KIRZ pseudogene from 2DL2-containing haplotypes (47). In this sense, 2DL2 can be considered an allele of the whole 2DL3-KIRZ-2DL1-gene region. The plasticity and rapid evolution of the KIR-gene family can have the effect of blurring distinctions between alleles and loci, as illustrated by the examples cited here. However, these complications and the questions they raise can all be successfully resolved by performing selected analysis and comparison at the level of the KIR haplotype.

Allelic Polymorphism Allelic polymorphism, as well as haplotypic difference in gene number and content, is also a significant component of human KIR diversity (43, 54, 57–60). Most KIR genes, like their Ly49 analogues (28), show allelic polymorphism, but the extent varies from one KIR gene to another. Most individuals are heterozygous at one or more KIR genes, and consequently the frequency with which unrelated humans have the same KIR type is very low (54, 57–59), perhaps approaching that of HLA. Allelic diversity for KIR genes is much higher than for CD94 and the NKG2 family of genes, in terms of both the number of alleles and the differences between them (61). Allelic diversity is generated by point mutation and homologous recombination (54, 57–60), but the relative contribution of these two mechanisms varies between the KIR genes (54). Allelic polymorphism has been described for all the

8 Feb 2002

8:43

AR

AR152-09.tex

AR152-09.SGM

LaTeX2e(2001/05/10)

P1: GJC

Annu. Rev. Immunol. 2002.20:217-251. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

KIR DIVERSITY

225

KIR genes encoding inhibitory HLA class I specific receptors, and it appears most extensive for 3DL1 and 3DL2 (54, 62, 63). The number of differences between KIR alleles is less than for polymorphic HLA class I loci. This property and the relatively small number of allelic sequences determined have limited the application of statistical methods in assessment of a role for natural selection in generating diversity. Although amino acid residues distinguishing KIR allotypes are distributed throughout the primary structure, their frequent localization in loops of the Ig-like domains is consistent with selection of polymorphisms that contribute to direct interaction with ligands or other molecules (54, 64, 65). However, the functional consequences of KIR polymorphism, for example, in modulating the strength or specificity of binding to MHC class I ligands have yet to be determined.

PHYLOGENETIC COMPARISON OF KIR The gene and allele diversity in human KIR haplotypes suggest that the KIRgene family has evolved rapidly in comparison to most human genes (25). Also pointing to this conclusion is the distribution of Alu repetitive elements in KIR haplotypes, which reveal a recent expansion of the gene family (38, 47). This expansion could have occurred subsequently to the divergence of primates from other mammalian orders. Comparison of KIR in several primate species (common chimpanzee, pygmy chimpanzee, rhesus monkey) has clearly demonstrated dramatic change in the KIR-gene family over time periods as short as a few million years (48, 61, 66). The most extensive comparison is that of humans and common chimpanzee (61), and it is used here to describe principles emerging from all such comparisons. Three KIR lineages are conserved in humans and chimpanzees, showing that the lineages predate divergence of the two species from a common ancestor (Figure 2) (61). The three lineages comprise: 2DL4 and 2DL5 receptors with D0-D2 domains (lineage 1); KIR3D specific for MHC-A and B allotypes (lineage 2); KIR recognizing MHC-C allotypes (lineage 3). In addition, the divergent KIRC1 is also found in the two species (61). Lineage 1 is the most conserved phylogenetically; 2DL4 and 2DL5 constitute two of the three orthologous genes in these species. Conservation of 2DL4 correlates with trans-species conservation of HLA-G (67), its ligand in humans (68, 69). In addition, the putative signaling motifs of 2DL4 provide an example of convergent evolution. In humans and both chimpanzee species, but not macaques (66), the cytoplasmic tail of 2DL4 has a single ITIM, the second one having lost its critical tyrosine through mutation. This property, combined with the presence of an arginine in the transmembrane region, could prevent SHP-1 binding and render 2DL4 an activating receptor (44). Surprisingly, different ITIMs are mutated in human (N-terminal) and chimpanzee (C-terminal) 2DL4 (48, 61, 70, 71). Furthermore, a common human 2DL4 variant has a short cytoplasmic tail with no ITIMs due to a frameshift (72). Thus, the three mutations were independently selected: either to evolve a new, and possibly

8 Feb 2002

8:43

Annu. Rev. Immunol. 2002.20:217-251. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

226

AR

AR152-09.tex

VILCHES

¥

AR152-09.SGM

LaTeX2e(2001/05/10)

P1: GJC

PARHAM

similar, function for human and chimpanzee 2DL4, or to eliminate a function that was no longer useful. The second lineage shared by humans and common chimpanzees comprises KIR3D specific for MHC-A and B. Humans have two KIR of this type, 3DL1 and 3DL2, whereas chimpanzees have a single receptor that shares sequence motifs with both (thus named 3DL1/2) and recognizes MHC-A and B allotypes of the two species (61). KIR recognizing MHC-C in humans (2DL1–3) and common chimpanzees (2DL6, 3DL4) belong to the third lineage, which is apparently specific to hominoids (Figure 2), as is the MHC-C gene (73, 74). This heterogeneous KIR lineage includes both KIR3D and type 1 KIR2D, the latter having evolved from a KIR3D ancestor by functional inactivation of the exon encoding the D0 domain (39, 40). The final stage of this evolutionary trend is seen in the human species, where all KIR recognizing HLA-C and their paralogs have nonfunctional pseudoexons 3 (39); the only remnant of the KIR3D ancestor is the KIR48/KIRX gene fragment (Figure 2). In the common chimpanzee, this lineage has two KIR with pseudoexons 3 (including 2DS4, the single other ortholog of a human KIR gene) and several genes encoding KIR3D (61). Loss of the D0 domain could relate to the lower expression of HLA-C than of HLA-A or -B (75, 76), and the loss could give more sensitive detection of cells in which HLA-C has undergone selective downregulation (39). Thus, both HLA-C and its specific receptors could have co-evolved to achieve a specialized function in NK-cell-mediated defense. In summary, comparison of KIR recognizing MHC-A, -B, and -C molecules between humans and common chimpanzees reveals conservation of functions, which contrasts with extensive species-specific evolution of the receptors exerting these functions. The mechanisms responsible for this divergence, which include gene duplication, recombination, point mutation, and inactivation of functional regions, have operated rapidly, perhaps at rates comparable to or even greater than associated with MHC class I genes (61). Extending the comparison of KIR genes to other primates (Figure 2) reveals that only 2DL4 has orthologs in all species studied; the phylogenetic relationship of rhesus 2DL5 to its homonyms with D0-D2 domains of other primates is less well supported. Thus, rhesus 2DL5 may represent a paralog of hominoid 2DL5. Of note, MHC-G is a pseudogene in the rhesus monkey (79, 80), but the Mamu-AG gene of this species, related genetically to its MHC-A, has structural features and expression patterns similar to HLA-G (81), making it of interest in determining whether Mamu-AG is a ligand for 2DL4. Genes encoding KIR with 3 Ig-domains, though diversified in several lineages (Figure 2), are also found in all species. Thus, from the known specificities of these two KIR families in humans and common chimpanzees, separate recognition of classical and nonclassical MHC molecules seems fundamental in the KIR system. Bonobos exemplify this best, since some animals have minimal haplotypes containing just 2DL4 and a KIR3D gene besides KIRC1 (48). It is also a constant that KIR3D and type 1 KIR2D are more diverse than KIR with D0-D2 domains, as are their respective ligands, which suggests a

8 Feb 2002

8:43

AR

AR152-09.tex

AR152-09.SGM

LaTeX2e(2001/05/10)

P1: GJC

Annu. Rev. Immunol. 2002.20:217-251. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

KIR DIVERSITY

227

cause-and-effect relationship between the polymorphisms of KIR and MHC, or the existence of a common force driving the evolution of both. KIR with similar extracellular domains but opposite signaling motifs coexist in all primate species studied (48, 61, 66), further supporting the hypothesis that activating KIR have important, but as yet unknown, roles in NK-cell function. The genes encoding activating KIR have expanded and diversified to greatest extent in humans (20, 21, 82, 83). Despite the diversity of KIR haplotypes, all humans have at least one gene for an activating KIR (51), with some donors bearing up to six receptors of this kind. In contrast, some common and pygmy chimpanzees lack short-tailed KIR without obvious impairment of their health and fecundity (48, 61). This could mean that activating KIR exert a redundant or accessory function in these species but, also, that they serve for defense against pathogens not encountered in captivity, or that additional activating KIR await discovery in these animals.

STRUCTURAL BASIS OF KIR-HLA RECOGNITION The structural heterogeneity of KIR and their specificity for polymorphic determinants of HLA class I molecules were established from study of NK cell cytotoxicity. In these analyses the susceptibility of allogeneic targets, and, later on, HLA class I– deficient cell lines transfected with single class I alleles, were tested against NKcell clones and KIR-transfected cells (25, 84–87). More recently, chimeras containing the extracellular portion of KIR and the activating cytoplasmic tails of the FcεRI γ -chain (88) or CD3ζ (89) have been used to overcome the technical difficulty of studying inhibitory signals. Further refinement in analysis of KIR-HLA interaction was achieved by synthesizing soluble forms of the receptor or its ligand. First, fusion proteins comprising the extracellular portion of KIR and the Fc of human IgG1 were used in flow cytometry to study cells expressing different HLA class I molecules (87, 90, 91). Direct interaction between recombinant KIR and HLA molecules was subsequently demonstrated by native gel electrophoresis, which revealed the basic 1:1 stoichiometry of KIR-HLA complexes (92). This technique was also used to study the specificity of KIR-HLA binding and the influence on it of other molecules (e.g., bivalent cations, antigenic peptides) (92–95). More recently, surface plasmon resonance (SPR) methodology has permitted quantitative assessment of KIRHLA interactions (96–99). The kinetics and thermodynamics of their binding have been measured by SPR, as well as the specificity and peptide-dependence of the interaction. The affinity of KIR for HLA-C is in the range shown by TCR for specific peptide-MHC complexes, but the kinetics of KIR-HLA-C binding and detachment are much faster. The fast on and off rates may facilitate progressive NK cell surveillance of MHC class I expression on a succession of potential target cells. Most recently, X-ray diffraction studies of crystallized 2DL2-Cw3 and 2DL1Cw4 complexes (65, 98), preceded by others on crystals of 2DL1–3 alone (64, 100,

8 Feb 2002

8:43

Annu. Rev. Immunol. 2002.20:217-251. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

228

AR

AR152-09.tex

VILCHES

¥

AR152-09.SGM

LaTeX2e(2001/05/10)

P1: GJC

PARHAM

101), have given three-dimensional pictures of HLA-C being recognized by inhibitory KIR. In several aspects of their three-dimensional structure KIR2DL1-3 resemble hemopoietic receptors. Most importantly, an acute angle between the D1 and D2 domains creates an elbow that constitutes the interface with HLA-C. This angle ranges from 55◦ to 84◦ , depending on the KIR and the method used to interpret raw data (64, 65, 98, 100, 101). Two loops of D1, three of D2, and the loop connecting the two Ig-like domains contribute to the interaction. All six loops bear negatively charged glutamate or aspartate residues that face a positively charged surface on HLA-C (Figure 3). The part of the HLA-C surface that becomes buried by interaction with KIR2DL1–2 includes the C-terminal end of the alpha-1 domain helix, the N-terminal end of the alpha-2 domain helix, and the C-terminal residues of the peptide. The shape complementarity of KIR-HLA-C surfaces, as assessed by the median shape correlation statistics value, is similar or greater than those of antigen-antibody or MHC-TCR complexes (65).

Recognition of HLA-C by KIR2DL1–3 MOLECULAR BASIS Much of the research to understand KIR-HLA interaction has concentrated on inhibitory KIR that recognize HLA-C. The specific recognition of HLA-C alleles with either lysine or asparagine in position 80 by 2DL1 and 2DL2-3 (84–86, 102), respectively, has been basically confirmed by studies with recombinant receptors and ligands (87, 90–92, 96, 103, 104). Nonetheless the latter studies have repeatedly shown weak to moderate cross-reactivity of 2DL1–3 with HLA-C alleles bearing the “wrong” epitope, an effect that depends partially on bound peptides (93, 97, 104). A few of these cross-reactions have been confirmed by cytotoxicity studies (104), but their relevance in vivo is unclear. The opposite phenomenon, failures in detecting expected interactions (such as that of Cw8 or Cw1 with 2DL2 and 2DL3), are less consistent and, in part, seem the result of a lower sensitivity of flow cytometry with KIR-Fc constructs in comparison with cytotoxicity assays (90, 104).

−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−− → Figure 3 (a) Amino acid sequence alignment of the regions participating in contacts in the 2DL2:Cw∗ 0304 and 2DL1:Cw∗ 0401 complexes (65, 98). The basis for the locus specificity of KIR2D is illustrated by the amino acid sequences of a typical Bw6 allele (B∗ 0702), the recombinant HLA-C/HLA-B allele B∗ 4601 and A∗ 0201. Contact amino acids in HLA-C and KIR2D are shaded. (b) Conserved and variable contacts in the KIR:HLA-C interface. Residues of Cw∗ 0304:GAVDPLLAL and Cw∗ 0401:QYDDAVYKL that participate in contacts with 2DL1 or 2DL2 are paired with their contacting amino acid(s). KIR residues are labeled according to the type of interaction they establish: salt bridges (bold), hydrogen bonds (underlined) or hydrophobic contacts (italics); other polar interactions (65) are not shown; amino acids implicated in more than one type of interaction are labeled with a combination of styles; for simplification, all HLA-C amino acids are in plain style. Residues differing between 2DL1 and 2DL2 are marked with asterisks.

Annu. Rev. Immunol. 2002.20:217-251. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

8 Feb 2002 8:43 AR AR152-09.tex AR152-09.SGM LaTeX2e(2001/05/10) P1: GJC

KIR DIVERSITY

229

8 Feb 2002

8:43

Annu. Rev. Immunol. 2002.20:217-251. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

230

AR

AR152-09.tex

VILCHES

¥

AR152-09.SGM

LaTeX2e(2001/05/10)

P1: GJC

PARHAM

Site-directed mutagenesis experiments and crystallographic analysis have clarified how KIR2D recognize HLA-C allotypes. Allospecificity is mostly determined by dimorphisms of KIR-residue 44 (2DL1Met44/2DL2–3Lys44) and amino acid 80 of HLA-C (Lys/Asn) (103–105), which establish direct interactions with each other (Figure 3) (65, 98). All other KIR-contact amino acids of Cw3 and Cw4 are completely conserved among HLA-C allotypes. However, the binding of 2DL2 to Cw3 and 2DL1 to Cw4 are based on largely different interactions (Figure 3a), which result in Lys80 of Cw4 being accommodated by a cavity of 2DL1 that is not seen in 2DL2 (65, 98). The nature of KIR-HLA-C contacts explains the profound effects of mutations affecting KIR2DL1-3 residues 44, 45, 68, 70, 105, 106, 135 and 183 (89, 91, 98, 103, 104). HLA-C residue 77, for which the Ser/Asn dimorphism is in strong linkage disequilibrium with that of residue 80, does not contribute directly to KIR recognition, as expected from the results of mutagenesis experiments (105). In contrast, the suggested influence of two additional HLA-C dimorphisms (Ala/Thr73 and Ala/Asp90) in the strength of binding to 2DL1–3 (106) cannot be explained by the crystallographic data. Furthermore, Cw∗ 1503, one of few natural alleles encoding the Ala73-Ala90 combination suggested to abolish KIR interaction (106), is recognized by both 2DL1-Fc and 2DL2-Fc constructs (104). Further research should establish definitively whether polymorphic motifs other than Asn/Lys80 can modulate recognition of HLA-C alleles by KIR2D. KIR recognizing HLA-CAsn80 do not cross-react with Bw6-positive HLA-B allotypes, although they have the same amino acid sequence at positions 77–80 as the HLA-C ligands (104). The locus-specificity of 2DL2 and -3 is determined, at least partly, by amino acid Val76 of HLA-C (Glu in HLA-B) because these KIR recognize B∗ 4601, an exceptional HLA-B allotype carrying Val76 (104, 107) (Figure 3a). Since Val76 and neighboring amino acids establish several hydrophobic contacts, the negative charge of HLA-B Glu76 would tend to disrupt these and thus change or destabilize the structure. Arginine 69 of Cw3 can also contribute to locus-specific recognition because it establishes hydrophilic interactions with Glu21 of 2DL2; HLA-B allotypes have nonpolar amino acids (Ala or Thr) at position 69 (Figure 3b). Lack of KIR2D cross-reactivity with HLA-A is explained by the fact that all allotypes of this locus bear three or more nonconservative changes in the contact residues: positions 69, 76, 79, 80, 145, and 149–151 (Figure 3a) (108). THE ROLE OF PEPTIDE The observation that certain antigenic peptides favor or prevent KIR recognition of HLA class I (86) raised the possibility that NK cells might discriminate between cells presenting self and foreign peptides. More detailed analysis showed that KIR recognition is indeed influenced by the peptide bound to HLA, but that there is no truly peptide-specific recognition, since KIR can bind to a same HLA molecule carrying rather different peptides (109–111). According to experiments with synthetic peptides, residue Ä–1 (P8 in nonamers) is most critical for recognition of HLA-B and HLA-C (98, 109, 110, 112–114). In the 2DL2:Cw3 complex, steric hindrance limits the peptide repertoire to amino acids with small side chains at position Ä–1 (98). In the 2DL1:Cw4 interface, P8

8 Feb 2002

8:43

AR

AR152-09.tex

AR152-09.SGM

LaTeX2e(2001/05/10)

P1: GJC

Annu. Rev. Immunol. 2002.20:217-251. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

KIR DIVERSITY

231

forms no contact with the KIR; the electronegative surface of 2DL1 seems to repel peptides with acidic residues in that position (65). Of note, the influence of peptides on KIR interaction varies for different HLA-C molecules (65, 93, 98), which may contribute to the variable recognition of allotypes sharing the same nominal specificity according to the residue at position 80 (104). The concept that emerges from the cited experiments is that peptides can be either permissive or prohibitive to KIR recognition of HLA. Despite the lack of peptide specificity in the KIR:HLA interaction, it is still relevant to ask whether peptides can influence NK cell recognition of infected cells. A priori it might seem unlikely that peptides provided by an intracellular pathogen are predominantly of a type that prevent recognition of HLA molecules. However, transfection of a cell line with a plasmid vector skewed the profile of peptides bound by HLA-B to an extent that prevented engagement of KIR3DL1, thus making the cells susceptible to lysis by NK cells expressing this KIR (114). One should perhaps wonder whether the basal set of peptides presented by healthy cells contains a proportion of permissive peptides tuned just above the threshold that permits inhibition of NK cells; even subtle changes in that pool might then trigger NK cell activation. This could be particularly relevant for HLA-C molecules, which are expressed at one tenth to one third the level of HLA-A and B (75, 76, 115) and are therefore closer to the threshold necessary for NK-cell inhibition. FROM HLA-C RECOGNITION TO NK CELL INHIBITION. THE ROLE OF BIVALENT CATIONS AND KIR AGGREGATION The mechanism by which HLA-C binding triggers phos-

phorylation of KIR ITIMs has yet to be identified. The angle between the D1 and D2 domains of 2DL1–2 seems to vary slightly upon HLA-C binding [1◦ –11◦ depending on the report (64, 65, 98, 100)]. Unknown is whether this variation induces a conformational change in the cytoplasmic tail that facilitates tyrosine phosphorylation. There is more experimental evidence for HLA-C inducing a KIR-aggregation process mediated by bivalent metallic cations: ■

■

■

■

Clustering of KIR on cell surfaces follows recognition of a protective HLAC molecule on a target cell, a process that is blocked by addition of zincchelating agents (116). Inhibitory function of 2DL1 is impaired by mutation at a Zn-binding motif of its N-terminal end or by addition of Zn-chelating agents, but its binding to HLA-C is unaffected by either manipulation (88). Zn2+ and other bivalent cations, including Cu2+ and Co2+, induce homo- or hetero-aggregation of recombinant KIR2D (95, 117). Soluble dimers of KIR2DL1, either engineered or Co2+-induced, bind more strongly to HLA-C than the monomer, which could be due either to higher affinity or avidity caused by proximity of twin binding sites (94, 95).

These findings provide evidence for KIR2D aggregation in the cell membrane after recognition of HLA-C. Two types of KIR2D homoaggregates have been

8 Feb 2002

8:43

Annu. Rev. Immunol. 2002.20:217-251. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

232

AR

AR152-09.tex

VILCHES

¥

AR152-09.SGM

LaTeX2e(2001/05/10)

P1: GJC

PARHAM

proposed as possible signaling mediators: binding of two molecules through their D2 domains, which would bring together their cytoplasmic tails; and higher order aggregates in which the N-terminal ends would be linked by a bivalent cation (95). The first model is appealing because of the structural similarity of KIR to hematopoietic receptors (100), which themselves form homo- or heterodimers. However, although crystallography has revealed KIR-KIR and KIR-HLA-C interactions different from the one believed to be physiological in ligand binding (64, 65, 98), none of them resembles the proposed aggregates nor seems capable of contributing to a higher affinity receptor or bringing together the cytoplasmic tails of several KIR. On the other hand, since KIR crystals were formed in the absence of zinc, or other bivalent cations, complexes of KIR bound through their N-terminal end were unlikely to have been seen. The Zn-binding capacity of KIR2D was inferred from the presence of an HEGVH motif in the N-terminal sequence (118). This sequence fits the zincbinding motif of neutral metallopeptidases, in which the two histidines coordinate with zinc and the glutamate is an essential part of the catalytic site (PROSITEDatabase of protein families and domains: http://www.expasy.ch/cgi-bin/nicedoc. pl?PDOC00129). In contrast, only His1 of KIR seems essential for Zn-binding, consistent with the cation bridging two KIR molecules through their N-terminal histidines (88, 95). Although Glu2 would in this model be unnecessary for metalinduced homodimerization (95), it is conserved in all inhibitory human and chimpanzee KIR with D1-D2 configuration, but not in KIR3D (43, 61). Further exploration is needed to see if Glu2 is necessary in other steps of KIR2D function. For example, mutation in Glu2 might reveal unsuspected interactions or even show that KIR2D can enzymatically degrade proteins involved in regulating NK-cell activity.

Specificity of Activating KIR2D The binding of HLA-C to KIR2DS1 and -2, although not quantified, is clearly weaker than to their inhibitory counterparts 2DL1–3 (91, 104, 119). Nonetheless, binding specificity of KIR2DS1 for Lys80 HLA-C and KIR2DS2 for Asn80 HLAC allotypes could be reproduced in flow cytometric assays using Fc constructs of 2DS1 (91) and a 2DS2Y45F mutant with enhanced affinity (104). All 2DL1 residues that contact Cw∗ 0401 are conserved in 2DS1 (Figure 4), but the nonconservative substitution of Lys for Thr70 seems to determine the lower affinity of the latter receptor (91). Similarly, only a Phe-to-Tyr change distinguishes the interacting loops of 2DL2 from those of 2DS2, but swapping this residue enhances the affinity of 2DS2 for HLA-CAsn80 allotypes (104, 119). Among the other human KIR2D with activating character, 2DS3 and 2DS5 resemble 2DL1 at residues implicated in HLA-C recognition, whereas 2DS4 is more similar to 2DL2 (Figure 4). The actual specificities of 2DS3 and 2DS5 have yet to be investigated, and as their extracellular domains bear one or more substitutions in the loops implicated in HLA-C recognition, reliable prediction

8 Feb 2002

8:43

AR

AR152-09.tex

AR152-09.SGM

LaTeX2e(2001/05/10)

P1: GJC

Annu. Rev. Immunol. 2002.20:217-251. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

KIR DIVERSITY

233

Figure 4 Activating KIR2D bear nonconservative amino acid substitutions for HLA-C-contacting residues of inhibitory KIR2D (shaded ).

of their specificity is not possible. As for 2DS4, some reports show recognition of certain, but not all, HLA-C allotypes (120–122), while other studies failed to demonstrate interaction with any HLA-A, -B or -C allotypes (93, 104, 119, 123). The existence of more KIR2DS receptors than known ligands challenges current understanding of KIR:HLA interactions and requires further investigation on the specificity and function of these KIR.

The Recognition of HLA-B and HLA-A by KIR3D KIR3DL1 was described as a specific receptor for HLA-B allotypes expressing the serological Bw4 epitope (124–129). This epitope, encoded by amino acids 80–83 in the highly polymorphic alpha-helix of the alpha-1 domain (130), is polymorphic itself: Three sequences (IALR, TALR, TLLR) have been found in positions 80–83 of different HLA-B allotypes recognized by anti-Bw4 sera (Figure 5). (A fourth

Figure 5 Comparison of HLA-B allotypes expressing the Bw4 epitope with HLA-A and HLA-B allotypes not recognized by KIR3DL1. The alignment shows the alpha helices in the regions homologous to those recognized by KIR2D in HLA-C molecules (contact positions of the latter are shaded).

8 Feb 2002

8:43

Annu. Rev. Immunol. 2002.20:217-251. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

234

AR

AR152-09.tex

VILCHES

¥

AR152-09.SGM

LaTeX2e(2001/05/10)

P1: GJC

PARHAM

Bw4 motif, TAAR, only found in the now deleted B∗ 4401 allotype proved to be the result of sequencing error (131). HLA-B alleles encoding the alternative serological Bw6 epitope and the majority of HLA-A molecules have NLRG and TLRG sequences, respectively, in positions 80–83. However, the HLA-A antigens A23 (9), 24 (9), 25 (10), and 32 (19) have a Bw4-associated motif (IALR) (Figure 5) and cross-react with anti-Bw4 antibodies (132). Specific NK-cell recognition of cells expressing the Bw4-positive HLA-B allotypes was demonstrated and shown to correlate with staining by the DX9 and Z27 antibodies, which turned out to recognize KIR3DL1 (124–129, 133). Early reports suggested that NK cells discriminate not only between Bw4 and Bw6 targets, but also among Bw4 alleles having either Thr80 or Ile80 (134, 135); also some NK cells recognize both HLA-A and HLA-B alleles with the IALR motif (134). Unfortunately, possible participation of KIR3DL1 or other MHC-receptors in NK-cell inhibition was not tested in these experiments, which have not been reproduced. KIR3DL1 discrimination of Bw4- and Bw6-allotypes was confirmed subsequently and shown to depend on amino acids Thr/Ile80, Leu82, and Arg83 by site-directed mutagenesis (57, 136, 137). In contrast with previous studies (134), HLA-A allotypes with the IALR motif did not inhibit NK cells expressing 3DL1 (136), suggesting that recognition of additional polymorphic residues outside the 80–83 region of HLA-B (Figure 5) confers locus-specificity to this KIR. As is the case for KIR2D:HLA-C complexes, the interaction of 3DL1 with HLA-B was susceptible to changes in the antigenic peptide repertoire (112–114). On the receptor side, study of deletion mutants containing only D0, D0 + D1, and D1 + D2 domains demonstrated requirement for all three Ig-domains of KIR3DL1 (133). In the context of the three-dimensional structure of KIR2D (64, 65, 98, 100, 101), the dependence of HLA-B recognition on the 3DL1 D0 is intriguing because the structure provides no clues regarding the role of this domain, even if it is assumed that the D1-D2 domains of KIR3D and KIR2D adopt similar spatial conformations. In comparison with the KIR2D:HLA-C interaction, knowledge of KIR3D:HLA-B binding has benefited less from the new methods for studying molecular interactions. Further study is needed to define which 3DL1 residues are essential for Bw4 recognition and how the polymorphisms of the many 3DL1 variants so far identified influence its function (54, 62, 63). The Ig-like domains of 3DS1, the short-tailed allotype of 3DL1, differ by only 6–12 amino acid changes from their 3DL1 counterparts. However, most of them are in loops of the Ig-like domains (54), including ones corresponding to the binding site of KIR2D (Figure 6), and they could therefore affect the binding properties of 3DS1. Indeed, no interaction of 3DS1 with Bw4 HLA-B alleles has been reported. Identification of KIR recognizing HLA-C and HLA-B prompted search for a specific HLA-A receptor. Independent investigations reported KIR3DL2 to be such a receptor. In one study, NK-cell clones expressing 3DL2 were inhibited by targets expressing HLA-A3 and A11, whereas targets expressing HLA-A1 or A2 were killed (129). In another study HLA-A3 was again shown to be a 3DL2 ligand, both through inhibition of cytotoxicity and using a KIR-Fc construct. In

8 Feb 2002

8:43

AR

AR152-09.tex

AR152-09.SGM

LaTeX2e(2001/05/10)

P1: GJC

Annu. Rev. Immunol. 2002.20:217-251. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

KIR DIVERSITY

235

Figure 6 Amino acid sequence alignment of human KIR3DL1 (L41269), KIRDS1 (L76661), and KIR3DL2 (L41270), and common chimpanzee KIR3DL1/2 (AF258798) in the regions homologous to the ligand-interacting loops of 2DL1 and 2DL2. Numbering refers to human KIR3D.

contrast with the first study, HLA-A11 seemed not to be recognized by 3DL2, and neither were A2, A68(28), A24(9), A31(19), and A33(19) (138). An inhibitory effect of HLA-A3 compared to -A2 mirrored the results of earlier work performed before identification of killer-cell receptors for HLA (139, 140). Although some investigations failed to observe recognition of HLA-A3 by 3DL2 (57), further evidence that HLA-A allotypes are recognized by KIR3D came from study of the common chimpanzee receptor KIR3DL1/2. As mentioned above, this receptor shares sequence motifs with the extracellular domains of human 3DL1and 3DL2 (Figure 6) and recognizes MHC-A, as well as MHC-B molecules of both humans and apes (61). Inconsistency in demonstrating 3DL2:HLA-A interactions might reflect a lower affinity than those other KIR:HLA pairs. In any case, the actual ligand of KIR3DL2 is currently a matter of controversy that needs to be resolved.

The Recognition of HLA-G by KIR2DL4 Inhibition of NK cells by HLA-G presents an attractive hypothesis to explain maternal tolerance of the trophoblast, a tissue that lacks HLA-A and HLA-B: expressing only HLA-C and the nonclassical class I molecules HLA-G and HLAE (141–143). There is agreement that HLA-G can directly inhibit NK cells as a ligand for the ILT-2 receptor (23, 144, 145) and through recognition of its leader peptide in complex with HLA-E by CD94:NKG2A (17–19). By contrast, the role of KIR in recognizing HLA-G is confused, with virtually every inhibitory KIR having been at some time proposed to be an HLA-G receptor (120, 146, 147). On the basis of the evidence currently available the best candidate is KIR2DL4 (69, 71, 148). In flow cytometry HLA-G transfected cells exhibited moderately brighter staining with a 2DL4-Fc construct than did other HLA class I transfectants (71). A more specific binding of 2DL4-Fc to HLA-G was shown by Rajagopalan & Long but “nonspecific binding” to other HLA class I molecules was also consistently detected (69). These authors reasoned that the D0 domain of 2DL4 was responsible for the nonspecific binding, because other constructs containing D0 domains had similar nonspecific binding (69). Contrasting with these results, an HLA-G tetramer

8 Feb 2002

8:43

Annu. Rev. Immunol. 2002.20:217-251. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

236

AR

AR152-09.tex

VILCHES

¥

AR152-09.SGM

LaTeX2e(2001/05/10)

P1: GJC

PARHAM

failed to bind to NK cells (144), which ubiquitously express 2DL4. Of several interpretations of this result, one possibility is that 2DL4 protein is not present on the cell surface of peripheral blood NK cells as reported in one study (148), but not in another (69). KIR2DL4 was proposed to mediate HLA-G-induced inhibition of decidual NK-cells, since protection by HLA-G was abrogated by an antiserum raised against 2DL4 (148). NK92 cells infected with vaccinia-2DL4 also appeared to be inhibited by HLA-G, but these experiments were later shown to lack a negative control, since mock infection produced occasionally the same effect (69). A further complication is the suggestion that 2DL4 activates, rather than inhibits, NK cells (44). Recognition of nonclassical, rather than classical MHC class I molecules is consistent with 2DL4’s divergent structure compared to KIR recognizing HLA-A, -B, and -C (70, 71). Unlike these KIR, 2DL4 is phylogenetically conserved in humans and Old World primates (Figure 3), fitting well with recognition of conserved, nonpolymorphic class I. However, what remains obscure is the physiological function of interaction between HLA-G and 2DL4. The recently reported KIR2DL5 gene is structurally related to 2DL4 (46) and similarly conserved in primates (Figure 3). These similarities raise the possibility that 2DL5 also recognizes a nonclassical MHC class I molecule. That the amino acid sequences of 2DL4 and 2DL5 are only 80% identical suggests they have different ligands (46).

THE EXPRESSION OF KIR GENES A diverse NK cell repertoire in which each cell expresses a subset of KIR is itself evidence for a model in which these receptors survey for abnormal and selective downregulation of HLA class I allotypes. With few exceptions, all KIR genes in the genome of every individual are transcribed in his/her polyclonal NK cell population (51). However, individual NK-cell clones express only some of the KIR genes, in apparently stochastic combinations that are regulated mainly at the transcriptional level (57) and stably maintained (140, 149, 150). Randomness in the activation of KIR genes is supported by their combinatorial frequencies (57), which fit a “product rule” (151). Similarly for alleles, individual NK cells from the same donor have been shown to express one, two, or no KIR3DL1 alleles, there being no allelic exclusion (54). Correlation of NK cell patterns of KIR expression with KIR and HLA genotype in siblings shows that the KIR repertoire is largely determined by the KIR genotype and that the modifying effect of HLA genotype is comparatively small (H. G. Shilling, L. A. Guethlein, N. W. Cheng, C. M. Gardiner, D. Tyan, and P. Parham, submitted for publication). This situation contrasts with the αβ T cell receptor repertoire, which is strongly biased as a result of thymic selection by autologous MHC type. The apparent weakness of the selection can in large part be attributed to the fact that CD94:NKG2A can also serve as an inhibitory receptor

8 Feb 2002

8:43

AR

AR152-09.tex

AR152-09.SGM

LaTeX2e(2001/05/10)

P1: GJC

Annu. Rev. Immunol. 2002.20:217-251. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

KIR DIVERSITY

237

for self-MHC class I. The evidence suggests CD94:NKG2A expression is being regulated so that it fills the holes in the KIR repertoire, i.e., it is expressed on those cells that do not express an inhibitory KIR with specificity for autologous HLA class I. Fitting with this interpretation is the inverse correlation seen between the proportions of total NK cells expressing KIR or CD94 (57; H. G. Shilling, L. A. Guethlein, N. W. Cheng, C. M. Gardiner, D. Tyan, and P. Parham, submitted for publication). Common principles seem to apply to the expression of most KIR genes, regardless of whether they encode inhibitory or activating receptors. KIR2DL4 is exceptional in being expressed ubiquitously, whereas all other KIR are expressed in clonotypic manner (57, 68, 69); controversial is whether the 2DL4 protein is also expressed ubiquitously (69) or restricted to decidual NK cells (68). Dissociation between transcription and protein expression has occasionally been reported for other KIR genes (152). There are also exceptions to the rule that KIR genes are expressed by at least some NK cells in every donor who has the gene: 2DL5 is silent in some individuals (42) and KIRC1 is generally not transcribed (45, 46), although a cDNA sequence was recently reported (GenBank AF352324). Patterns of KIR gene expression correlate with sequence variation in their promoter regions (42, 47), and the lack of expression of some genes appears intrinsic, not a result of downregulation. KIR-gene expression in T lymphocytes and NK cells seems governed by similar rules. One key difference is that T-lymphocytes acquire KIR after having participated in an immune response; activation of KIR genes takes place after rearrangement of the Tcr genes, possibly during transition to becoming memory cells, a process to which KIR may contribute (152–156). Since KIR inhibit the effector functions of T-cells (157), an important issue is whether and how memory T-lymphocytes are activated by a renewed contact with antigen after they express KIR. Requirement for a strong stimulus to overcome KIR-mediated inhibition seems antagonistic to effective memory response. Whereas KIR expression appears to be as stable in T-lymphocytes as in NK cells (149, 150, 154), one report described progressive downregulation of T-cell KIR in the absence of stimulation through the TCR (34). The molecular mechanisms that regulate expression and cell distribution of KIR remain unknown. Comparison of the promoter sequences of transcribed and silent variants of the KIR2DL5 gene points to consensus binding sites for AML1 and Ets-1 as necessary for KIR transcription (42). In addition, TCF-1 has been proposed to participate in the stochastic activation of some Ly49 genes (158); the core sequence of the TCF-1 binding site [(C/A)A(C/A)AG] appears in the promoter of several ILT (159) and KIR genes (42), making possible the participation of TCF-1 in their regulation. Comparative promoter analysis of expressed and silent KIR genes and variants should identify which cis- and trans-acting elements are essential for stochastic and tissue-specific activation of KIR genes. Because of the convergence of Ly49 and KIR to analogous functions and mode of expression, despite considerable structural difference (151), parallel study of both systems is

8 Feb 2002

8:43

238

AR

AR152-09.tex

VILCHES

¥

AR152-09.SGM

LaTeX2e(2001/05/10)

P1: GJC

PARHAM

essential for appreciation of the mechanisms that have produced such surprising convergence.

Annu. Rev. Immunol. 2002.20:217-251. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

KIR DIVERSITY: WHAT IS IT GOOD FOR? The variability of the KIR system in human populations raises the obvious question: What is it good for? The most extreme aspect of KIR variability is that particular genes can be absent from a person’s genome. In actual fact, only 3DL2, 2DL4, and KIRC1 are ubiquitously present in KIR haplotypes and are thus framework genes of the cluster (49). Of these genes, KIRC1 is silent in most humans (45, 46). For 2DL4, there is uncertainty regarding its signaling potential (44, 69, 72) and even on its membrane expression in nondecidual NK cells (68). Furthermore, an adult fertile woman appears to lack a 2DL4 gene (N. G´omez-Lozano, R. de Pablo, C. Vilches, unpublished), and embryos homozygous for HLA-G1 defects develop normally (160), although products derived from alternatively processed mRNA of these mutants have been suggested to interact with 2DL4 (161). The single conserved inhibitory KIR could thus be 3DL2, one for which most humans seem not to have a ligand. If the polymorphism of both the HLA system and the KIR-gene complex are considered together, inactivation of KIR:HLA pairs appears a common natural phenomenon (Table 3). However, although no particular KIR seems essential for survival until reproductive age, no deficiencies affecting the expression of all KIR genes have been reported. It is unlikely that having more KIR genes confers a

TABLE 3 Natural “knock-outs” of KIR genes and KIR-HLA pairs KIR:HLA pair

Inactivating event

2DL1:HLA-CLys80

HLA-CAsn80 homozygotes (common) Lack of 2DL1 gene (uncommon)

2DL2,3:HLA-CAsn80

HLA-CLys80 homozygotes (common) No known examples of negatives for both 2DL2 and 2DL3

3DL1:HLA-BBw4

Bw6 homozygotes (common) 3DS1 homozygotes (uncommon)

3DL2:HLA-A3,others?

Combinations of other HLA-A alleles (most humans) No known examples of negatives for 3DL2

2DL4:HLA-G

2DL4 variants without known signaling motifs (common) HLA-G mutations (rare) Lack of 2DL4 gene (rare)

2DL5:unknown ligand

Lack of the gene or lack of expression (common)

3DS1, 2DS1-5:HLA-C or unknown ligands

Lack of individual genes (common) Mutation of DAP12 (rare)

8 Feb 2002

8:43

AR

AR152-09.tex

AR152-09.SGM

LaTeX2e(2001/05/10)

P1: GJC

Annu. Rev. Immunol. 2002.20:217-251. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

KIR DIVERSITY

239

strong evolutionary advantage in the long term, since haplotypes bearing fewer KIR genes are as frequent as those with a higher number (51). However, the presence of different numbers and qualities of KIR genes in the genome likely influences the NK-cell mediated immune response. For instance, four inhibitory KIR:HLA pairs would function in individuals having a KIR ‘A’ haplotype whose HLA phenotype included A3, a Bw4+ HLA-B allotype, and both Asn80 and Lys80 in their HLA-C allotypes. On the contrary, donors homozygous for the KIR ‘B’ haplotype shown in Table 2 could have no single functional inhibitory KIR:HLA pair if they were homozygous for HLAC-Lys80 and lacked HLA-A3. These people could still use the CD94:NKG2A system for sensing levels of HLA class I expression, but it is conceivable that their NK cells would be less able to defend against pathogens inducing selective downregulation of certain HLA molecules, or that the use of different receptors confers qualitative differences to their response against pathogens. It has been argued in the latter direction that the faster progression of HIV infection in B35 individuals may be related to the Bw6 condition of this antigen (31); the rationale for this proposal is that the downregulation of HLA-B by HIVNef would not be detected by NK cells of an individual who is homozygous for Bw6. More recently, Bw6 homozygosity has indeed been proposed to correlate with a prognosis for a person with HIV infection (162). In comparison to other species, an intriguing aspect of the human KIR cluster is the expansion of genes encoding activating KIR; some donors have up to six genes of this type (A/B genotypes). As discussed above, convergent evolution has produced multiple activating versions of KIR and other MHC class I receptors, which suggests they have an essential, but as yet unknown, role. Activating KIR have been suggested to participate in the lysis of HSV-infected cells (36) and in the pathogenesis of rheumatoid arthritis (163, 164). Paradoxically, no obvious impairment in number or phenotype of NK-cells or T-cells has been observed in patients homozygous for mutations affecting the DAP12 gene (165). In contrast, both NK-cell phenotype and function, and T-lymphocyte-mediated, antigen-specific responses, are altered in DAP12-targeted mice (166). Furthermore, the activating Ly49H receptor confers resistance to potentially lethal infection by murine CMV (167, 168), which constitutes the strongest evidence for the relevance of activating MHC class I– specific receptors in immunity. Activating KIR might participate in NK-cell function in two ways: first, by being truly specific for HLA; second, by recognizing HLA class I-cross-reactive molecules encoded, induced, or modified by pathogens. Some viral proteins indeed mimic MHC class I (169) or mask cellular MHC class I–homologues that activate NK cells (9). Similar proteins could be recognized by activating KIR and still be beneficial to the pathogen by favoring a balance that permits the host to survive and, hence, the virus to continue its life cycle. If, on the other hand, the more parsimonious possibility is assumed of activating KIR being specific for HLA class I, these receptors could complement the role of inhibitory KIR in surveillance against selective downregulation of MHC class I; altered cells that lose some

8 Feb 2002

8:43

Annu. Rev. Immunol. 2002.20:217-251. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

240

AR

AR152-09.tex

VILCHES

¥

AR152-09.SGM

LaTeX2e(2001/05/10)

P1: GJC

PARHAM

fraction of their HLA class I molecules could activate NK cells co-expressing inhibitor and activating receptors with different HLA class I specificities even in the absence of any other marker of transformation. Since T-lymphocytes express inhibitory and activating KIR (120, 157, 163, 164, 170) and NK cells secrete immunomodulatory factors (2), antigen-specific immune responses could vary among humans bearing dissimilar KIR receptors. KIR might achieve this by raising or lowering the activation threshold of T-lymphocytes, promoting termination or continuation of immune responses, favoring or hindering presentation of antigens by particular HLA loci due to competition with TCR for overlapping surfaces of MHC, thereby skewing decision-making (tolerance/activation, effector/memory/apoptosis, Th1/Th2) by T-lymphocytes. In summary, KIR diversity has the potential to contribute variability in both innate and adaptive immunity. Availability of simple methods to study KIR genes makes it likely that the influence of their diversity on the immune response will be analyzed in the short term by means of epidemiological studies. These should illustrate whether KIR polymorphisms behave as factors of susceptibility or protection that influence the response to infections, malignancy, autoimmune and inflammatory diseases, and transplanted tissue. ACKNOWLEDGMENT The unpublished research by N. G´omez-Lozano and C. Vilches that is referred to here is supported by grant FIS 01/0381 from the Instituto de Salud Carlos III, Spain. Visit the Annual Reviews home page at www.annualreviews.org

LITERATURE CITED 1. Trinchieri G. 1989. Biology of natural killer cells. Adv. Immunol. 47:187–376 2. Biron CA, Nguyen KB, Pien GC, Cousens LP, Salazar-Mather TP. 1999. Natural killer cells in antiviral defense: function and regulation by innate cytokines. Annu. Rev. Immunol. 17:189–220 3. Kubin MZ, Parshley DL, Din W, Waugh JY, Davis-Smith T, Smith CA, Macduff BM, Armitage RJ, Chin W, Cassiano L, Borges L, Petersen M, Trinchieri G, Goodwin RG. 1999. Molecular cloning and biological characterization of NK cell activation-inducing ligand, a counterstructure for CD48. Eur. J. Immunol. 29:3466–77

4. Lanier LL. 2000. Turning on natural killer cells. J. Exp. Med. 191:1259–62 5. Moretta A, Biassoni R, Bottino C, Mingari MC, Moretta L. 2000. Natural cytotoxicity receptors that trigger human NKcell-mediated cytolysis. Immunol. Today 21:228–34 6. Tangye SG, Cherwinski H, Lanier LL, Phillips JH. 2000. 2B4-mediated activation of human natural killer cells. Mol. Immunol. 37:493–501 7. Lanier LL. 2001. On guard-activating NK cell receptors. Nat. Immunol. 2:23–27 8. Moretta A, Bottino C, Vitale M, Pende D, Cantoni C, Mingari MC, Biassoni R, Moretta L. 2001. Activating receptors

8 Feb 2002

8:43

AR

AR152-09.tex

AR152-09.SGM

LaTeX2e(2001/05/10)

P1: GJC

KIR DIVERSITY

Annu. Rev. Immunol. 2002.20:217-251. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

9.

10.

11.

12. 13.

14.

15.

16.

17.

18.

and coreceptors involved in human natural killer cell-mediated cytolysis. Annu. Rev. Immunol. 19:197–223 Cosman D, Mullberg J, Sutherland CL, Chin W, Armitage R, Fanslow W, Kubin M, Chalupny NJ. 2001. ULBPs, novel MHC class I-related molecules, bind to CMV glycoprotein UL16 and stimulate NK cytotoxicity through the NKG2D receptor. Immunity 14:123–33 Kubin M, Cassiano L, Chalupny J, Chin W, Cosman D, Fanslow W, M¨ullberg J, Rousseau AM, Ulrich D, Armitage R. 2001. ULBP1, 2, 3: novel MHC class I-related molecules that bind to human cytomegalovirus glycoprotein UL16, activate NK cells. Eur. J. Immunol. 31:1428– 37 Karre K, Ljunggren HG, Piontek G, Kiessling R. 1986. Selective rejection of H-2-deficient lymphoma variants suggests alternative immune defence strategy. Nature 319:675–78 Karre K. 1997. How to recognize a foreign submarine. Immunol. Rev. 155:5–9 Brodsky FM, Lem L, Solache A, Bennett EM. 1999. Human pathogen subversion of antigen presentation. Immunol. Rev. 168:199–215 Farrell HE, Degli-Esposti MA, DavisPoynter NJ. 1999. Cytomegalovirus evasion of natural killer cell responses. Immunol. Rev. 168:187–97 Tortorella D, Gewurz BE, Furman MH, Schust DJ, Ploegh HL. 2000. Viral subversion of the immune system. Annu. Rev. Immunol. 18:861–926 Algarra I, Cabrera T, Garrido F. 2000. The HLA crossroad in tumor immunology. Hum. Immunol. 61:65–73 Lee N, Llano M, Carretero M, Ishitani A, Navarro F, Lopez-Botet M, Geraghty DE. 1998. HLA-E is a major ligand for the natural killer inhibitory receptor CD94/NKG2A. Proc. Natl. Acad. Sci. USA 95:5199–204 Braud VM, Allan DS, O’Callaghan CA, Soderstrom K, D’Andrea A, Ogg GS,

19.

20.

21.

22.

23.

24.

25.

241

Lazetic S, Young NT, Bell JI, Phillips JH, Lanier LL, McMichael AJ. 1998. HLA-E binds to natural killer cell receptors CD94/NKG2A, B and C. Nature 391:795–99 Borrego F, Ulbrecht M, Weiss EH, Coligan JE, Brooks AG. 1998. Recognition of human histocompatibility leukocyte antigen (HLA)-E complexed with HLA class I signal sequence-derived peptides by CD94/NKG2 confers protection from natural killer cell-mediated lysis. J. Exp. Med. 187:813–18 Wagtmann N, Biassoni R, Cantoni C, Verdiani S, Malnati MS, Vitale M, Bottino C, Moretta L, Moretta A, Long EO. 1995. Molecular clones of the p58 NK cell receptor reveal immunoglobulinrelated molecules with diversity in both the extra- and intracellular domains. Immunity 2:439–49 Colonna M, Samaridis J. 1995. Cloning of immunoglobulin-superfamily members associated with HLA-C and HLA-B recognition by human natural killer cells. Science 268:405–8 Borges L, Hsu ML, Fanger N, Kubin M, Cosman D. 1997. A family of human lymphoid and myeloid Ig-like receptors, some of which bind to MHC class I molecules. J. Immunol. 159:5192–96 Colonna M, Navarro F, Bellon T, Llano M, Garcia P, Samaridis J, Angman L, Cella M, Lopez-Botet M. 1997. A common inhibitory receptor for major histocompatibility complex class I molecules on human lymphoid and myelomonocytic cells. J. Exp. Med. 186:1809–18 Lee N, Goodlett DR, Ishitani A, Marquardt H, Geraghty DE. 1998. HLA-E surface expression depends on binding of TAP-dependent peptides derived from certain HLA class I signal sequences. J. Immunol. 160:4951–60 Valiante NM, Lienert K, Shilling HG, Smits BJ, Parham P. 1997. Killer cell receptors: keeping pace with MHC class I evolution. Immunol. Rev. 155:155–64

8 Feb 2002

8:43

Annu. Rev. Immunol. 2002.20:217-251. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

242

AR

AR152-09.tex

VILCHES

¥

AR152-09.SGM

LaTeX2e(2001/05/10)

P1: GJC

PARHAM

26. Vance RE, Kraft JR, Altman JD, Jensen PE, Raulet DH. 1998. Mouse CD94/NKG2A is a natural killer cell receptor for the nonclassical major histocompatibility complex (MHC) class I molecule Qa-1(b). J. Exp. Med. 188: 1841–48 27. Raulet DH, Held W, Correa I, Dorfman JR, Wu MF, Corral L. 1997. Specificity, tolerance and developmental regulation of natural killer cells defined by expression of class I-specific Ly49 receptors. Immunol. Rev. 155:41–52 28. Takei F, Brennan J, Mager DL. 1997. The Ly-49 family: genes, proteins and recognition of class I MHC. Immunol. Rev. 155:67–77 29. Hanke T, Takizawa H, McMahon CW, Busch DH, Pamer EG, Miller JD, Altman JD, Liu Y, Cado D, Lemonnier FA, Bjorkman PJ, Raulet DH. 1999. Direct assessment of MHC class I binding by seven Ly49 inhibitory NK cell receptors. Immunity 11:67–77 30. Tomasec P, Braud VM, Rickards C, Powell MB, McSharry BP, Gadola S, Cerundolo V, Borysiewicz LK, McMichael AJ, Wilkinson GW. 2000. Surface expression of HLA-E, an inhibitor of natural killer cells, enhanced by human cytomegalovirus gpUL40. Science 287:1031 31. Cohen GB, Gandhi RT, Davis DM, Mandelboim O, Chen BK, Strominger JL, Baltimore D. 1999. The selective downregulation of class I major histocompatibility complex proteins by HIV-1 protects HIV-infected cells from NK cells. Immunity 10:661–71 32. Le Gall S, Erdtmann L, Benichou S, Berlioz-Torrent C, Liu L, Benarous R, Heard JM, Schwartz O. 1998. Nef interacts with the mu subunit of clathrin adaptor complexes and reveals a cryptic sorting signal in MHC I molecules. Immunity 8:483–95 33. Schust DJ, Tortorella D, Seebach J, Phan C, Ploegh HL. 1998. Trophoblast class I

34.

35.

36.

37.

38.

39.

40.

41.

42.

major histocompatibility complex (MHC) products are resistant to rapid degradation imposed by the human cytomegalovirus (HCMV) gene products US2 and US11. J. Exp. Med. 188:497–503 Huard B, Fruh K. 2000. A role for MHC class I down-regulation in NK cell lysis of herpes virus-infected cells. Eur. J. Immunol. 30:509–15 L´opez-Botet M, Llano M, Ortega M. 2001. Human cytomegalovirus and natural killer-mediated surveillance of HLA class I expression: a paradigm of host-pathogen adaptation. Immunol. Rev. 181:193–202 Pietra G, Semino C, Cagnoni F, Boni L, Cangemi G, Frumento G, Melioli G. 2000. Natural killer cells lyse autologous herpes simplex virus infected targets using cytolytic mechanisms distributed clonotypically. J. Med. Virol. 62:354–63 Lanier LL, Corliss BC, Wu J, Leong C, Phillips JH. 1998. Immunoreceptor DAP12 bearing a tyrosine-based activation motif is involved in activating NK cells. Nature 391:703–7 Martin AM, Freitas EM, Witt CS, Christiansen FT. 2000. The genomic organization and evolution of the natural killer immunoglobulin-like receptor (KIR) gene cluster. Immunogenetics 51:268–80 Vilches C, Pando MJ, Parham P. 2000. Genes encoding human killer-cell Ig-like receptors with D1 and D2 extracellular domains all contain untranslated pseudoexons encoding a third Ig-like domain. Immunogenetics 51:639–46 Wilson MJ, Torkar M, Trowsdale J. 1997. Genomic organization of a human killer cell inhibitory receptor gene. Tissue Antigens 49:574–79 Selvakumar A, Steffens U, Palanisamy N, Chaganti RS, Dupont B. 1997. Genomic organization and allelic polymorphism of the human killer cell inhibitory receptor gene KIR103. Tissue Antigens 49:564–73 Vilches C, Gardiner CM, Parham P.

8 Feb 2002

8:43

AR

AR152-09.tex

AR152-09.SGM

LaTeX2e(2001/05/10)

P1: GJC

KIR DIVERSITY

43.

Annu. Rev. Immunol. 2002.20:217-251. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

44.

44a.

45.

46.

47.

47a.

48.

2000. Gene structure and promoter variation of expressed and non-expressed variants of the KIR2DL5 gene. J. Immunol. 165:6416–21 Steffens U, Vyas Y, Dupont B, Selvakumar A. 1998. Nucleotide and amino acid sequence alignment for human killer cell inhibitory receptors (KIR), 1998. Tissue Antigens 51:398–413 Marti F, Xu CW, Selvakumar A, Brent R, Dupont B, King PD. 1998. LCKphosphorylated human killer cell-inhibitory receptors recruit and activate phosphatidylinositol 3-kinase. Proc. Natl. Acad. Sci. USA 95:11810–15 Rajagopalan S, Fu J, Long EO. 2001. Induction of IFN-gamma production but not cytotoxicity by the killer cell Iglike receptor KIR2DL4 (CD158d) in resting NK cells. J. Immunol. 167:1877– 81 Torkar M, Norgate Z, Colonna M, Trowsdale J, Wilson MJ. 1998. Isotypic variation of novel immunoglobulin-like transcript/killer cell inhibitory receptor loci in the leukocyte receptor complex. Eur. J. Immunol. 28:3959–67 Vilches C, Rajalingam R, Uhrberg M, Gardiner CM, Young NT, Parham P. 2000. KIR2DL5, a novel killer-cell receptor with a D0-D2 configuration of Ig-like domains. J. Immunol. 164:5797–804 Wilson MJ, Torkar M, Haude A, Milne S, Jones T, Sheer D, Beck S, Trowsdale J. 2000. Plasticity in the organization and sequences of human KIR/ILT gene families. Proc. Natl. Acad. Sci. USA 97:4778– 83 Long EO, Barber DF, Burshtyn DN, Faure M, Peterson M, Rajagopalan S, Renard V, Sandusky M, Stebbins CC, Wagtmann N, Watzl C. 2001. Inhibition of natural killer cell activation signals by killer cell immunoglobulin-like receptors (CD158). Immunol. Rev. 181:223–33 Rajalingam R, Hong M, Adams EJ, Shum BP, Guethlein LA, Parham P. 2001. Short KIR haplotypes in pygmy chim-

49.

50.

51.

52.

53.

54.

55.

56.

243

panzee (Bonobo) resemble the conserved framework of diverse human KIR haplotypes. J. Exp. Med. 193:135–46 Barten R, Torkar M, Haude A, Trowsdale J, Wilson MJ. 2001. Divergent and convergent evolution of NK-cell receptors. Trends Immunol. 22:52–57 Wende H, Colonna M, Ziegler A, Volz A. 1999. Organization of the leukocyte receptor cluster (LRC) on human chromosome 19q13.4. Mamm. Genome 10:154– 60 Uhrberg M, Valiante NM, Shum BP, Shilling HG, Lienert-Weidenbach K, Corliss B, Tyan D, Lanier LL, Parham P. 1997. Human diversity in killer cell inhibitory receptor genes. Immunity 7:753– 63 Witt CS, Dewing C, Sayer DC, Uhrberg M, Parham P, Christiansen FT. 1999. Population frequencies and putative haplotypes of the killer cell immunoglobulinlike receptor sequences and evidence for recombination. Transplantation 68:1784– 89 Crum KA, Logue SE, Curran MD, Middleton D. 2000. Development of a PCRSSOP approach capable of defining the natural killer cell inhibitory receptor (KIR) gene sequence repertoires. Tissue Antigens 56:313–26 Gardiner CM, Guethlein LA, Shilling HG, Pando M, Carr WH, Rajalingam R, Vilches C, Parham P. 2001. Different NK cell surface phenotypes defined by the DX9 antibody are due to KIR3DL1 gene polymorphism. J. Immunol. 166:2992– 3001 Vilches C, Pando MJ, Rajalingam R, Gardiner CM, Parham P. 2000. Discovery of two novel variants of KIR2DS5 reveals this gene to be a common component of human KIR ‘B’ haplotypes. Tissue Antigens 56:453–56 G´omez-Lozano N, Gardiner CM, Parham P, Vilches C. 2001. Killer-cell immunoglobulin-like receptor gene haplotypes contain between zero and two

8 Feb 2002

8:43

244

57.

Annu. Rev. Immunol. 2002.20:217-251. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

58.

59.

60.

61.

62.

63.

64.

65.

AR

AR152-09.tex

VILCHES

¥

AR152-09.SGM

LaTeX2e(2001/05/10)

P1: GJC

PARHAM

KIR2DL5 loci. Eur. J. Immunogenet. 28: 284 (Abstr.) Valiante NM, Uhrberg M, Shilling HG, Lienert-Weidenbach K, Arnett KL, D’Andrea A, Phillips JH, Lanier LL, Parham P. 1997. Functionally and structurally distinct NK cell receptor repertoires in the peripheral blood of two human donors. Immunity 7:739–51 Shilling HG, Lienert-Weidenbach K, Valiante NM, Uhrberg M, Parham P. 1998. Evidence for recombination as a mechanism for KIR diversification. Immunogenetics 48:413–16 Rajalingam R, Gardiner CM, Canavez F, Vilches C, Parham P. 2001. Identification of seventeen novel KIR variants: fourteen of them from two non-Caucasian donors. Tissue Antigens 57:22–31 Chwae YJ, Cho SE, Kim SJ, Kim J. 1999. Diversity of the repertoire of p58 killer cell inhibitory receptors in a single individual. Immunol. Lett. 68:267–74 Khakoo SI, Rajalingam R, Shum BP, Weidenbach K, Flodin L, Muir DG, Canavez F, Cooper SL, Valiante NM, Lanier LL, Parham P. 2000. Rapid evolution of NK cell receptor systems demonstrated by comparison of chimpanzees and humans. Immunity 12:687–98 Vyas Y, Selvakumar A, Steffens U, Dupont B. 1998. Multiple transcripts of the killer cell immunoglobulin-like receptor family, KIR3DL1 (NKB1), are expressed by natural killer cells of a single individual. Tissue Antigens 52:510–19 Kwon D, Chwae YJ, Choi IH, Park JH, Kim SJ, Kim J. 2000. Diversity of the p70 killer cell inhibitory receptor (KIR3DL) family members in a single individual. Mol. Cells 10:54–60 Snyder GA, Brooks AG, Sun PD. 1999. Crystal structure of the HLA-Cw3 allotype-specific killer cell inhibitory receptor KIR2DL2. Proc. Natl. Acad. Sci. USA 96:3864–69 Fan QR, Long EO, Wiley DC. 2001. Crystal structure of the human natural

66.

67.

68.

69.

70.

71.

72.

73.

74.

killer cell inhibitory receptor KIR2DL1HLA-Cw4 complex. Nat. Immunol. 2:452–60 Hershberger KL, Shyam R, Miura A, Letvin NL. 2001. Diversity of the killer cell Ig-like receptors of rhesus monkeys. J. Immunol. 166:4380–90 Arnaiz-Villena A, Martinez-Laso J, Alvarez M, Castro MJ, Varela P, GomezCasado E, Suarez B, Recio MJ, VargasAlarcon G, Morales P. 1997. Primate Mhc-E and -G alleles. Immunogenetics 46:251–66 Cantoni C, Falco M, Pessino A, Moretta A, Moretta L, Biassoni R. 1999. P49, a putative HLA-G1 specific inhibitory NK receptor belonging to the immunoglobulin superfamily. J. Reprod. Immunol. 43:157–65 Rajagopalan S, Long EO. 1999. A human histocompatibility leukocyte antigen (HLA)-G-specific receptor expressed on all natural killer cells. J. Exp. Med. 189:1093–99. Correction. 2000. J. Exp. Med. 191:29 Selvakumar A, Steffens U, Dupont B. 1996. NK cell receptor gene of the KIR family with two IG domains but highest homology to KIR receptors with three IG domains. Tissue Antigens 48:285–94 Cantoni C, Verdiani S, Falco M, Pessino A, Cilli M, Conte R, Pende D, Ponte M, Mikaelsson MS, Moretta L, Biassoni R. 1998. p49, a putative HLA class I-specific inhibitory NK receptor belonging to the immunoglobulin superfamily. Eur. J. Immunol. 28:1980–90 Witt CS, Martin A, Christiansen FT. 2000. Detection of KIR2DL4 alleles by sequencing and SSCP reveals a common allele with a shortened cytoplasmic tail. Tissue Antigens 56:248–57 Adams EJ, Thomson G, Parham P. 1999. Evidence for an HLA-C-like locus in the orangutan Pongo pygmaeus. Immunogenetics 49:865–71 Vogel TU, Evans DT, Urvater JA, O’Connor DH, Hughes AL, Watkins DI. 1999.

8 Feb 2002

8:43

AR

AR152-09.tex

AR152-09.SGM

LaTeX2e(2001/05/10)

P1: GJC

KIR DIVERSITY

75.

Annu. Rev. Immunol. 2002.20:217-251. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

76.

77.

78.

79.

80.

81.

82.

83.

Major histocompatibility complex class I genes in primates: co-evolution with pathogens. Immunol. Rev. 167:327–37 McCutcheon JA, Gumperz J, Smith KD, Lutz CT, Parham P. 1995. Low HLA-C expression at cell surfaces correlates with increased turnover of heavy chain mRNA. J. Exp. Med. 181:2085–95 Neisig A, Melief CJ, Neefjes J. 1998. Reduced cell surface expression of HLA-C molecules correlates with restricted peptide binding and stable TAP interaction. J. Immunol. 160:171–79 Saitou N, Nei M. 1987. The neighbourjoining method: a new method for reconstructing phylogenetic trees. Mol. Biol. Evol. 4:406–25 Felsenstein J. 1985. Confidence limits on phylogenies: an approach using the bootstrap. Evolution 4:783–91 Castro MJ, Morales P, Fernandez-Soria V, Suarez B, Recio MJ, Alvarez M, Martin-Villa M, Arnaiz-Villena A. 1996. Allelic diversity at the primate Mhc-G locus: exon 3 bears stop codons in all Cercopithecinae sequences. Immunogenetics 43:327–36 Boyson JE, Iwanaga KK, Golos TG, Watkins DI. 1996. Identification of the rhesus monkey HLA-G ortholog. MamuG is a pseudogene. J. Immunol. 157:5428– 37 Boyson JE, Iwanaga KK, Golos TG, Watkins DI. 1997. Identification of a novel MHC class I gene, Mamu-AG, expressed in the placenta of a primate with an inactivated G locus. J. Immunol. 159:3311– 21 Dohring C, Samaridis J, Colonna M. 1996. Alternatively spliced forms of human killer inhibitory receptors. Immunogenetics 44:227–30 Biassoni R, Cantoni C, Falco M, Verdiani S, Bottino C, Vitale M, Conte R, Poggi A, Moretta A, Moretta L. 1996. The human leukocyte antigen (HLA)-C-specific “activatory” or “inhibitory” natural killer cell receptors display highly homologous

84.

85.

86.

87.

88.

89.

90.

91.

92.

245

extracellular domains but differ in their transmembrane and intracytoplasmic portions. J. Exp. Med. 183:645–50 Moretta A, Biassoni R, Bottino C, Pende D, Vitale M, Poggi A, Mingari MC, Moretta L. 1997. Major histocompatibility complex class I-specific receptors on human natural killer and T lymphocytes. Immunol. Rev. 155:105–17 Colonna M. 1997. Specificity and function of immunoglobulin superfamily NK cell inhibitory and stimulatory receptors. Immunol. Rev. 155:127–33 Long EO, Burshtyn DN, Clark WP, Peruzzi M, Rajagopalan S, Rojo S, Wagtmann N, Winter CC. 1997. Killer cell inhibitory receptors: diversity, specificity, and function. Immunol. Rev. 155:135–44 Wagtmann N, Rajagopalan S, Winter CC, Peruzzi M, Long EO. 1995. Killer cell inhibitory receptors specific for HLA-C and HLA-B identified by direct binding and by functional transfer. Immunity 3:801–9 Rajagopalan S, Long EO. 1998. Zinc bound to the killer cell-inhibitory receptor modulates the negative signal in human NK cells. J. Immunol. 161:1299–305 Richardson J, Reyburn HT, Luque I, Vales-Gomez M, Strominger JL. 2000. Definition of polymorphic residues on killer Ig-like receptor proteins which contribute to the HLA-C binding site. Eur. J. Immunol. 30:1480–85 Dohring C, Colonna M. 1996. Human natural killer cell inhibitory receptors bind to HLA class I molecules. Eur. J. Immunol. 26:365–69 Biassoni R, Pessino A, Malaspina A, Cantoni C, Bottino C, Sivori S, Moretta L, Moretta A. 1997. Role of amino acid position 70 in the binding affinity of p50.1 and p58.1 receptors for HLA-Cw4 molecules. Eur. J. Immunol. 27:3095–99 Fan QR, Garboczi DN, Winter CC, Wagtmann N, Long EO, Wiley DC. 1996. Direct binding of a soluble natural killer cell inhibitory receptor to a soluble human leukocyte antigen-Cw4 class I major

8 Feb 2002

8:43

246

93.

Annu. Rev. Immunol. 2002.20:217-251. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

94.

95.

96.

97.

98.

99.

100.

101.

AR

AR152-09.tex

VILCHES

¥

AR152-09.SGM

LaTeX2e(2001/05/10)

P1: GJC

PARHAM

histocompatibility complex molecule. Proc. Natl. Acad. Sci. USA 93:7178–83 Kim J, Chwae YJ, Kim MY, Choi IH, Park JH, Kim SJ. 1997. Molecular basis of HLA-C recognition by p58 natural killer cell inhibitory receptors. J. Immunol. 159:3875–82 Fan QR, Long EO, Wiley DC. 2000. A disulfide-linked natural killer cell receptor dimer has higher affinity for HLA-C than wild-type monomer. Eur. J. Immunol. 30:2692–97 Fan QR, Long EO, Wiley DC. 2000. Cobalt-mediated dimerization of the human natural killer cell inhibitory receptor. J. Biol. Chem. 275:23,700–6 Vales-Gomez M, Reyburn HT, Mandelboim M, Strominger JL. 1998. Kinetics of interaction of HLA-C ligands with natural killer cell inhibitory receptors. Immunity 9:337–44 Maenaka K, Juji T, Nakayama T, Wyer JR, Gao GF, Maenaka T, Zaccai NR, Kikuchi A, Yabe T, Tokunaga K, Tadokoro K, Stuart DI, Jones EY, van der Merwe PA. 1999. Killer cell immunoglobulin receptors and T cell receptors bind peptide-major histocompatibility complex class I with distinct thermodynamic and kinetic properties. J. Biol. Chem. 274:28329–34 Boyington JC, Motyka SA, Schuck P, Brooks AG, Sun PD. 2000. Crystal structure of an NK cell immunoglobulin-like receptor in complex with its class I MHC ligand. Nature 405:537–43 Vales-Gomez M, Reyburn H, Strominger J. 2000. Molecular analyses of the interactions between human NK receptors and their HLA ligands. Hum. Immunol. 61:28–38 Fan QR, Mosyak L, Winter CC, Wagtmann N, Long EO, Wiley DC. 1997. Structure of the inhibitory receptor for human natural killer cells resembles haematopoietic receptors. Nature 389:96– 100 Maenaka K, Juji T, Stuart DI, Jones EY.

102.

103.

104.

105.

106.

107.

108.

1999. Crystal structure of the human p58 killer cell inhibitory receptor (KIR2DL3) specific for HLA-Cw3-related MHC class I. Structure 7:391–98 Reyburn H, Mandelboim O, Vales-Gomez M, Sheu EG, Pazmany L, Davis DM, Strominger JL. 1997. Human NK cells: their ligands, receptors and functions. Immunol. Rev. 155:119–25 Winter CC, Long EO. 1997. A single amino acid in the p58 killer cell inhibitory receptor controls the ability of natural killer cells to discriminate between the two groups of HLA-C allotypes. J. Immunol. 158:4026–28 Winter CC, Gumperz JE, Parham P, Long EO, Wagtmann N. 1998. Direct binding and functional transfer of NK cell inhibitory receptors reveal novel patterns of HLA-C allotype recognition. J. Immunol. 161:571–77 Mandelboim O, Reyburn HT, ValesGomez M, Pazmany L, Colonna M, Borsellino G, Strominger JL. 1996. Protection from lysis by natural killer cells of group 1 and 2 specificity is mediated by residue 80 in human histocompatibility leukocyte antigen C alleles and also occurs with empty major histocompatibility complex molecules. J. Exp. Med. 184:913–22 Mandelboim O, Reyburn HT, Sheu EG, Vales-Gomez M, Davis DM, Pazmany L, Strominger JL. 1997. The binding site of NK receptors on HLA-C molecules. Immunity 6:341–50 Barber LD, Percival L, Valiante NM, Chen L, Lee C, Gumperz JE, Phillips JH, Lanier LL, Bigge JC, Parekh RB, Parham P. 1996. The inter-locus recombinant HLA-B∗ 4601 has high selectivity in peptide binding and functions characteristic of HLA-C. J. Exp. Med. 184:735–40 Robinson J, Malik A, Parham P, Bodmer JG, Marsh SG. 2000. IMGT/HLA database—a sequence database for the human major histocompatibility complex. Tissue Antigens 55:280–87

8 Feb 2002

8:43

AR

AR152-09.tex

AR152-09.SGM

LaTeX2e(2001/05/10)

P1: GJC

Annu. Rev. Immunol. 2002.20:217-251. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

KIR DIVERSITY 109. Rajagopalan S, Long EO. 1997. The direct binding of a p58 killer cell inhibitory receptor to human histocompatibility leukocyte antigen (HLA)-Cw4 exhibits peptide selectivity. J. Exp. Med. 185:1523–28 110. Mandelboim O, Wilson SB, Vales-Gomez M, Reyburn HT, Strominger JL. 1997. Self and viral peptides can initiate lysis by autologous natural killer cells. Proc. Natl. Acad. Sci. USA 94:4604–9 111. Zappacosta F, Borrego F, Brooks AG, Parker KC, Coligan JE. 1997. Peptides isolated from HLA-Cw∗ 0304 confer different degrees of protection from natural killer cell-mediated lysis. Proc. Natl. Acad. Sci. USA 94:6313–18 112. Peruzzi M, Wagtmann N, Long EO. 1996. A p70 killer cell inhibitory receptor specific for several HLA-B allotypes discriminates among peptides bound to HLA-B∗ 2705. J. Exp. Med. 184:1585–90 113. Peruzzi M, Parker KC, Long EO, Malnati MS. 1996. Peptide sequence requirements for the recognition of HLA-B∗ 2705 by specific natural killer cells. J. Immunol. 157:3350–56 114. Liberatore C, Capanni M, Albi N, Volpi I, Urbani E, Ruggeri L, Mencarelli A, Grignani F, Velardi A. 1999. Natural killer cell-mediated lysis of autologous cells modified by gene therapy. J. Exp. Med. 189:1855–62 115. Snary D, Barnstable CJ, Bodmer WF, Crumpton MJ. 1977. Molecular structure of human histocompatibility antigens: the HLA-C series. Eur. J. Immunol. 7:580–85 116. Davis DM, Chiu I, Fassett M, Cohen GB, Mandelboim O, Strominger JL. 1999. The human natural killer cell immune synapse. Proc. Natl. Acad. Sci. USA 96:15062– 67 117. Vales-Gomez M, Erskine RA, Deacon MP, Strominger JL, Reyburn HT. 2001. The role of zinc in the binding of killer cell Ig-like receptors to class I MHC proteins. Proc. Natl. Acad. Sci. USA 98:1734–39 118. Rajagopalan S, Winter CC, Wagtmann

119.

120.

121.

122.

123.

124.

125.

247

N, Long EO. 1995. The Ig-related killer cell inhibitory receptor binds zinc and requires zinc for recognition of HLA-C on target cells. J. Immunol. 155:4143–46 Vales-Gomez M, Reyburn HT, Erskine RA, Strominger J. 1998. Differential binding to HLA-C of p50-activating and p58-inhibitory natural killer cell receptors. Proc. Natl. Acad. Sci. USA 95: 14,326–31 Mandelboim O, Davis DM, Reyburn HT, Vales-Gomez M, Sheu EG, Pazmany L, Strominger JL. 1996. Enhancement of class II-restricted T cell responses by costimulatory NK receptors for class I MHC proteins. Science 274:2097–100 Campbell KS, Cella M, Carretero M, Lopez-Botet M, Colonna M. 1998. Signaling through human killer cell activating receptors triggers tyrosine phosphorylation of an associated protein complex. Eur. J. Immunol. 28:599–609 Katz G, Markel G, Mizrahi S, Arnon TI, Mandelboim O. 2001. Recognition of HLA-Cw4 but not HLA-Cw6 by the NK cell receptor killer cell Ig-like receptor two-domain short tail number 4. J. Immunol. 166:7260–67 Bottino C, Sivori S, Vitale M, Cantoni C, Falco M, Pende D, Morelli L, Augugliaro R, Semenzato G, Biassoni R, Moretta L, Moretta A. 1996. A novel surface molecule homologous to the p58/p50 family of receptors is selectively expressed on a subset of human natural killer cells and induces both triggering of cell functions and proliferation. Eur. J. Immunol. 26:1816–24 Litwin V, Gumperz J, Parham P, Phillips JH, Lanier LL. 1994. NKB1: a natural killer cell receptor involved in the recognition of polymorphic HLA-B molecules. J. Exp. Med. 180:537–43 Lanier LL, Gumperz JE, Parham P, Melero I, Lopez-Botet M, Phillips JH. 1995. The NKB1 and HP-3E4 NK cells receptors are structurally distinct glycoproteins and independently recognize

8 Feb 2002

8:43

248

126.

Annu. Rev. Immunol. 2002.20:217-251. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

127.

128.

129.

130.

131.

132.

AR

AR152-09.tex

VILCHES

¥

AR152-09.SGM

LaTeX2e(2001/05/10)

P1: GJC

PARHAM

polymorphic HLA-B and HLA-C molecules. J. Immunol. 154:3320–27 Gumperz JE, Litwin V, Phillips JH, Lanier LL, Parham P. 1995. The Bw4 public epitope of HLA-B molecules confers reactivity with natural killer cell clones that express NKB1, a putative HLA receptor. J. Exp. Med. 181:1133–44 D’Andrea A, Chang C, Franz-Bacon K, McClanahan T, Phillips JH, Lanier LL. 1995. Molecular cloning of NKB1. A natural killer cell receptor for HLA-B allotypes. J. Immunol. 155:2306–10 Vitale M, Sivori S, Pende D, Augugliaro R, Di Donato C, Amoroso A, Malnati M, Bottino C, Moretta L, Moretta A. 1996. Physical and functional independency of p70 and p58 natural killer (NK) cell receptors for HLA class I: their role in the definition of different groups of alloreactive NK cell clones. Proc. Natl. Acad. Sci. USA 93:1453–57 Pende D, Biassoni R, Cantoni C, Verdiani S, Falco M, di Donato C, Accame L, Bottino C, Moretta A, Moretta L. 1996. The natural killer cell receptor specific for HLA-A allotypes: a novel member of the p58/p70 family of inhibitory receptors that is characterized by three immunoglobulin-like domains and is expressed as a 140-kD disulphide-linked dimer. J. Exp. Med. 184:505–18 Lutz CT, Smith KD, Greazel NS, Mace BE, Jensen DA, McCutcheon JA, Goeken NE. 1994. Bw4-reactive and Bw6reactive antibodies recognize multiple distinct HLA structures that partially overlap in the alpha-1 helix. J. Immunol. 153:4099–110 Bodmer JG, Marsh SG, Albert ED, Bodmer WF, Dupont B, Erlich HA, Mach B, Mayr WR, Parham P, Sasazuki T. 1994. Nomenclature for factors of the HLA system, 1994. Tissue Antigens 44:1– 18 Bodmer JG, Marsh SG, Albert ED, Bodmer WF, Bontrop RE, Charron D, Dupont B, Erlich HA, Fauchet R, Mach

133.

134.

135.

136.

137.

138.

139.

140.

B, Mayr WR, Parham P, Sasazuki T, Schreuder GM, Strominger JL, Svejgaard A, Terasaki PI. 1997. Nomenclature for factors of the HLA system, 1996. Tissue Antigens 49:297–321 Rojo S, Wagtmann N, Long EO. 1997. Binding of a soluble p70 killer cell inhibitory receptor to HLA-B∗ 5101: requirement for all three p70 immunoglobulin domains. Eur. J. Immunol. 27:568–71 Cella M, Longo A, Ferrara GB, Strominger JL, Colonna M. 1994. NK3specific natural killer cells are selectively inhibited by Bw4-positive HLA alleles with isoleucine 80. J. Exp. Med. 180:1235–42 Luque I, Solana R, Galiani MD, Gonz´alez R, Garcia F, L´opez de Castro JA, Pe˜na J. 1996. Threonine 80 on HLA-B27 confers protection against lysis by a group of natural killer clones. Eur. J. Immunol. 26:1974–77 Gumperz JE, Barber LD, Valiante NM, Percival L, Phillips JH, Lanier LL, Parham P. 1997. Conserved and variable residues within the Bw4 motif of HLAB make separable contributions to recognition by the NKB1 killer cell-inhibitory receptor. J. Immunol. 158:5237–41 Kurago ZB, Lutz CT, Smith KD, Colonna M. 1998. NK cell natural cytotoxicity and IFN-gamma production are not always coordinately regulated: engagement of DX9 KIR+ NK cells by HLAB7 variants and target cells. J. Immunol. 160:1573–80 Dohring C, Scheidegger D, Samaridis J, Cella M, Colonna M. 1996. A human killer inhibitory receptor specific for HLA-A. J. Immunol. 156:3098–101 Storkus WJ, Salter RD, Alexander J, Ward FE, Ruiz RE, Cresswell P, Dawson JR. 1991. Class I-induced resistance to natural killing: identification of nonpermissive residues in HLA-A2. Proc. Natl. Acad. Sci. USA 88:5989–92 Litwin V, Gumperz J, Parham P, Phillips JH, Lanier LL. 1993. Specificity of HLA

8 Feb 2002

8:43

AR

AR152-09.tex

AR152-09.SGM

LaTeX2e(2001/05/10)

P1: GJC

KIR DIVERSITY

Annu. Rev. Immunol. 2002.20:217-251. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

141.

142.

143.

144.

145.

146.

147.

class I antigen recognition by human NK clones: evidence for clonal heterogeneity, protection by self and non-self alleles, and influence of the target cell type. J. Exp. Med. 178:1321–36 King A, Hiby SE, Gardner L, Joseph S, Bowen JM, Verma S, Burrows TD, Loke YW. 2000. Recognition of trophoblast HLA class I molecules by decidual NK cell receptors—a review. Placenta 21 (Suppl. A): S81–85 King A, Burrows TD, Hiby SE, Bowen JM, Joseph S, Verma S, Lim PB, Gardner L, Le Bouteiller P, Ziegler A, UchanskaZiegler B, Loke YW. 2000. Surface expression of HLA-C antigen by human extravillous trophoblast. Placenta 21: 376–87 King A, Allan DS, Bowen M, Powis SJ, Joseph S, Verma S, Hiby SE, McMichael AJ, Loke YW, Braud VM. 2000. HLA-E is expressed on trophoblast and interacts with CD94/NKG2 receptors on decidual NK cells. Eur. J. Immunol. 30:1623– 31 Allan DS, Colonna M, Lanier LL, Churakova TD, Abrams JS, Ellis SA, McMichael AJ, Braud VM. 1999. Tetrameric complexes of human histocompatibility leukocyte antigen (HLA)-G bind to peripheral blood myelomonocytic cells. J. Exp. Med. 189:1149–56 Navarro F, Llano M, Bellon T, Colonna M, Geraghty DE, Lopez-Botet M. 1999. The ILT2(LIR1) and CD94/NKG2A NK cell receptors respectively recognize HLA-G1 and HLA-E molecules coexpressed on target cells. Eur. J. Immunol. 29:277–83 Mandelboim O, Pazmany L, Davis DM, Vales-Gomez M, Reyburn HT, Rybalov B, Strominger JL. 1997. Multiple receptors for HLA-G on human natural killer cells. Proc. Natl. Acad. Sci. USA 94: 14666–70 Munz C, Holmes N, King A, Loke YW, Colonna M, Schild H, Rammensee HG. 1997. Human histocompatibility leuko-

148.

149.

150.

151.

152.

153.

249

cyte antigen (HLA)-G molecules inhibit NKAT3 expressing natural killer cells. J. Exp. Med. 185:385–91 Ponte M, Cantoni C, Biassoni R, TradoriCappai A, Bentivoglio G, Vitale C, Bertone S, Moretta A, Moretta L, Mingari MC. 1999. Inhibitory receptors sensing HLA-G1 molecules in pregnancy: decidua-associated natural killer cells express LIR-1 and CD94/NKG2A and acquire p49, an HLA-G1-specific receptor. Proc. Natl. Acad. Sci. USA 96:5674– 79 Moretta A, Bottino C, Pende D, Tripodi G, Tambussi G, Viale O, Orengo A, Barbaresi M, Merli A, Ciccone E, et al. 1990. Identification of four subsets of human CD3-CD16+ natural killer (NK) cells by the expression of clonally distributed functional surface molecules: correlation between subset assignment of NK clones and ability to mediate specific alloantigen recognition. J. Exp. Med. 172:1589– 98 Moretta A, Tambussi G, Bottino C, Tripodi G, Merli A, Ciccone E, Pantaleo G, Moretta L. 1990. A novel surface antigen expressed by a subset of human CD3CD16+ natural killer cells. Role in cell activation and regulation of cytolytic function. J. Exp. Med. 171:695–714 Raulet DH, Vance RE, McMahon CW. 2001. Regulation of the natural killer cell receptor repertoire. Annu. Rev. Immunol. 19:291–330 Vely F, Peyrat M, Couedel C, Morcet J, Halary F, Davodeau F, Romagne F, Scotet E, Saulquin X, Houssaint E, Schleinitz N, Moretta A, Vivier E, Bonneville M. 2001. Regulation of inhibitory and activating killer-cell Ig-like receptor expression occurs in T cells after termination of TCR rearrangements. J. Immunol. 166:2487–94 Mingari MC, Schiavetti F, Ponte M, Vitale C, Maggi E, Romagnani S, Demarest J, Pantaleo G, Fauci AS, Moretta L. 1996. Human CD8+ T lymphocyte subsets that

8 Feb 2002

8:43

250

Annu. Rev. Immunol. 2002.20:217-251. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

154.

155.

156.

157.

158.

159.

160.

161.

AR

AR152-09.tex

VILCHES

¥

AR152-09.SGM

LaTeX2e(2001/05/10)

P1: GJC

PARHAM

express HLA class I-specific inhibitory receptors represent oligoclonally or monoclonally expanded cell populations. Proc. Natl. Acad. Sci. USA 93:12,433– 38 Uhrberg M, Valiante NM, Young NT, Lanier LL, Phillips JH, Parham P. 2001. The repertoire of killer-cell Ig-like receptor and CD94:NKG2A receptors in T cells: clones sharing identical alphabeta TCR rearrangement express highly diverse killer cell Ig-like receptor patterns. J. Immunol. 166:3923–32 Young NT, Uhrberg M, Phillips JH, Lanier LL, Parham P. 2001. Differential expression of leukocyte receptor complex-encoded Ig-like receptors correlates with the transition from effector to memory CTL. J. Immunol. 166:3933–41 Ugolini S, Arpin C, Anfossi N, Walzer T, Cambiaggi A, Forster R, Lipp M, Toes RE, Melief CJ, Marvel J, Vivier E. 2001. Involvement of inhibitory NKRs in the survival of a subset of memory-phenotype CD8+ T cells. Nat. Immunol. 2:430–35 Phillips JH, Gumperz JE, Parham P, Lanier LL. 1995. Superantigen-dependent, cell-mediated cytotoxicity inhibited by MHC class I receptors on T lymphocytes. Science 268:403–5 Held W, Kunz B, Lowin-Kropf B, van de Wetering M, Clevers H. 1999. Clonal acquisition of the Ly49A NK cell receptor is dependent on the trans-acting factor TCF-1. Immunity 11:433–42 Young NT, Canavez F, Uhrberg M, Shum BP, Parham P. 2001. Conserved organization of the ILT/LIR gene family within the polymorphic human leukocyte receptor complex. Immunogenetics. In press Ober C, Aldrich C, Rosinsky B, Robertson A, Walker MA, Willadsen S, Verp MS, Geraghty DE, Hunt JS. 1998. HLAG1 protein expression is not essential for fetal survival. Placenta 19:127–32 Riteau B, Rouas-Freiss N, Menier C, Paul P, Dausset J, Carosella ED. 2001. HLA-G2, -G3, and -G4 isoforms ex-

162.

163.

164.

165.

166.

167.

168.

pressed as nonmature cell surface glycoproteins inhibit NK and antigen-specific CTL cytolysis. J. Immunol. 166:5018– 26 Flores-Villanueva PO, Yunis EJ, Delgado JC, Vittinghoff E, Buchbinder S, Leung JY, Uglialoro AM, Clavijo OP, Rosenberg ES, Kalams SA, Braun JD, Boswell SL, Walker BD, Goldfeld AE. 2001. Control of HIV-1 viremia and protection from AIDS are associated with HLA-Bw4 homozygosity. Proc. Natl. Acad. Sci. USA 98:5140–45 Namekawa T, Snyder MR, Yen JH, Goehring BE, Leibson PJ, Weyand CM, Goronzy JJ. 2000. Killer cell activating receptors function as costimulatory molecules on CD4+CD28null T cells clonally expanded in rheumatoid arthritis. J. Immunol. 165:1138–45 Yen JH, Moore BE, Nakajima T, Scholl D, Schaid DJ, Weyand CM, Goronzy JJ. 2001. Major histocompatibility complex class I-recognizing receptors are disease risk genes in rheumatoid arthritis. J. Exp. Med. 193:1159–67 Paloneva J, Kestila M, Wu J, Salminen A, Bohling T, Ruotsalainen V, Hakola P, Bakker AB, Phillips JH, Pekkarinen P, Lanier LL, Timonen T, Peltonen L. 2000. Loss-of-function mutations in TYROBP (DAP12) result in a presenile dementia with bone cysts. Nat. Genet 25:357– 61 Bakker AB, Hoek RM, Cerwenka A, Blom B, Lucian L, McNeil T, Murray R, Phillips LH, Sedgwick JD, Lanier LL. 2000. DAP12-deficient mice fail to develop autoimmunity due to impaired antigen priming. Immunity 13:345–53 Brown MG, Dokun AO, Heusel JW, Smith HR, Beckman DL, Blattenberger EA, Dubbelde CE, Stone LR, Scalzo AA, Yokoyama WM. 2001. Vital involvement of a natural killer cell activation receptor in resistance to viral infection. Science 292:934–37 Lee SH, Girard S, Macina D, Busa M,

8 Feb 2002

8:43

AR

AR152-09.tex

AR152-09.SGM

LaTeX2e(2001/05/10)

P1: GJC

KIR DIVERSITY

Annu. Rev. Immunol. 2002.20:217-251. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

Zafer A, Belouchi A, Gros P, Vidal SM. 2001. Susceptibility to mouse cytomegalovirus is associated with deletion of an activating natural killer cell receptor of the C-type lectin superfamily. Nat. Genet. 28:42–45 169. Cosman D, Fanger N, Borges L. 1999. Human cytomegalovirus, MHC class I

251

and inhibitory signalling receptors: more questions than answers. Immunol. Rev. 168:177–85 170. Mingari MC, Moretta A, Moretta L. 1998. Regulation of KIR expression in human T cells: a safety mechanism that may impair protective T-cell responses. Immunol. Today 19:153–57

P1: FRK

February 1, 2002

12:47

Annual Reviews

AR152-FM

Annual Review of Immunology Volume 20, 2002

Annu. Rev. Immunol. 2002.20:217-251. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

CONTENTS FRONTISPIECE, Charles A. Janeway, Jr. A TRIP THROUGH MY LIFE WITH AN IMMUNOLOGICAL THEME, Charles A. Janeway, Jr.

xiv 1

THE B7 FAMILY OF LIGANDS AND ITS RECEPTORS: NEW PATHWAYS FOR COSTIMULATION AND INHIBITION OF IMMUNE RESPONSES, Beatriz M. Carreno and Mary Collins

29

MAP KINASES IN THE IMMUNE RESPONSE, Chen Dong, Roger J. Davis, and Richard A. Flavell

PROSPECTS FOR VACCINE PROTECTION AGAINST HIV-1 INFECTION AND AIDS, Norman L. Letvin, Dan H. Barouch, and David C. Montefiori T CELL RESPONSE IN EXPERIMENTAL AUTOIMMUNE ENCEPHALOMYELITIS (EAE): ROLE OF SELF AND CROSS-REACTIVE ANTIGENS IN SHAPING, TUNING, AND REGULATING THE AUTOPATHOGENIC T CELL REPERTOIRE, Vijay K. Kuchroo, Ana C. Anderson, Hanspeter Waldner, Markus Munder, Estelle Bettelli, and Lindsay B. Nicholson

55 73

101

NEUROENDOCRINE REGULATION OF IMMUNITY, Jeanette I. Webster, Leonardo Tonelli, and Esther M. Sternberg

125

MOLECULAR MECHANISM OF CLASS SWITCH RECOMBINATION: LINKAGE WITH SOMATIC HYPERMUTATION, Tasuku Honjo, Kazuo Kinoshita, and Masamichi Muramatsu

165

INNATE IMMUNE RECOGNITION, Charles A. Janeway, Jr. and Ruslan Medzhitov

KIR: DIVERSE, RAPIDLY EVOLVING RECEPTORS OF INNATE AND ADAPTIVE IMMUNITY, Carlos Vilches and Peter Parham ORIGINS AND FUNCTIONS OF B-1 CELLS WITH NOTES ON THE ROLE OF CD5, Robert Berland and Henry H. Wortis E PROTEIN FUNCTION IN LYMPHOCYTE DEVELOPMENT, Melanie W. Quong, William J. Romanow, and Cornelis Murre

197 217 253 301

LYMPHOCYTE-MEDIATED CYTOTOXICITY, John H. Russell and Timothy J. Ley

SIGNAL TRANSDUCTION MEDIATED BY THE T CELL ANTIGEN x

323

P1: FRK

Annu. Rev. Immunol. 2002.20:217-251. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

February 1, 2002

12:47

Annual Reviews

AR152-FM

CONTENTS

xi

RECEPTOR: THE ROLE OF ADAPTER PROTEINS, Lawrence E. Samelson INTERACTION OF HEAT SHOCK PROTEINS WITH PEPTIDES AND ANTIGEN PRESENTING CELLS: CHAPERONING OF THE INNATE AND ADAPTIVE IMMUNE RESPONSES, Pramod Srivastava CHROMATIN STRUCTURE AND GENE REGULATION IN THE IMMUNE SYSTEM, Stephen T. Smale and Amanda G. Fisher PRODUCING NATURE’S GENE-CHIPS: THE GENERATION OF PEPTIDES FOR DISPLAY BY MHC CLASS I MOLECULES, Nilabh Shastri, Susan

371

Schwab, and Thomas Serwold

395 427

463

THE IMMUNOLOGY OF MUCOSAL MODELS OF INFLAMMATION, Warren Strober, Ivan J. Fuss, and Richard S. Blumberg T CELL MEMORY, Jonathan Sprent and Charles D. Surh

GENETIC DISSECTION OF IMMUNITY TO MYCOBACTERIA: THE HUMAN MODEL, Jean-Laurent Casanova and Laurent Abel ANTIGEN PRESENTATION AND T CELL STIMULATION BY DENDRITIC CELLS, Pierre Guermonprez, Jenny Valladeau, Laurence Zitvogel, Clotilde Th´ery, and Sebastian Amigorena

495 551 581

621

NEGATIVE REGULATION OF IMMUNORECEPTOR SIGNALING, Andr´e Veillette, Sylvain Latour, and Dominique Davidson

669

CPG MOTIFS IN BACTERIAL DNA AND THEIR IMMUNE EFFECTS, Arthur M. Krieg

PROTEIN KINASE Cθ

709 IN

T CELL ACTIVATION, Noah Isakov and Amnon

Altman

RANK-L AND RANK: T CELLS, BONE LOSS, AND MAMMALIAN EVOLUTION, Lars E. Theill, William J. Boyle, and Josef M. Penninger PHAGOCYTOSIS OF MICROBES: COMPLEXITY IN ACTION, David M. Underhill and Adrian Ozinsky

761 795 825

STRUCTURE AND FUNCTION OF NATURAL KILLER CELL RECEPTORS: MULTIPLE MOLECULAR SOLUTIONS TO SELF, NONSELF DISCRIMINATION, Kannan Natarajan, Nazzareno Dimasi, Jian Wang, Roy A. Mariuzza, and David H. Margulies

853

INDEXES Subject Index Cumulative Index of Contributing Authors, Volumes 1–20 Cumulative Index of Chapter Titles, Volumes 1–20

ERRATA An online log of corrections to Annual Review of Immunology chapters (if any, 1997 to the present) may be found at http://immunol.annualreviews.org/

887 915 925

8 Feb 2002

9:44

AR

AR152-10.tex

AR152-10.SGM

LaTeX2e(2001/05/10) P1: GJC 10.1146/annurev.immunol.20.100301.064833

Annu. Rev. Immunol. 2002. 20:253–300 DOI: 10.1146/annurev.immunol.20.100301.064833 c 2002 by Annual Reviews. All rights reserved Copyright °

ORIGINS AND FUNCTIONS OF B-1 CELLS WITH NOTES ON THE ROLE OF CD5 Annu. Rev. Immunol. 2002.20:253-300. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

Robert Berland and Henry H. Wortis Department of Pathology, Tufts University School of Medicine and the Program in Immunology, Sackler School of Graduate Biomedical Sciences, Boston, Massachusetts 02111; e-mail: [email protected], [email protected]

Key Words autoimmunity, tolerance, repertoire, B cells, T independent type 2 (TI-2), antigen ■ Abstract Whether B-1a (CD5+) cells are a distinct lineage derived from committed fetal/neonatal precursors or arise from follicular B-2 cells in response to BCR ligation and other, unknown signals remains controversial. Recent evidence indicates that B-1a cells can derive from adult precursors expressing an appropriate specificity when the (self-) antigen is present. Antibody specificity determines whether a B cell expressing immunoglobulin transgenes has a B-2, B-1a or marginal zone (MZ) phenotype. MZ cells share many phenotypic characteristics of B-1 cells and, like them, appear to develop in response to T independent type 2 antigens. Because fetal-derived B cell progenitors fail to express terminal deoxynucleotidyl transferase (TdT) and for other reasons, they are likely to express a repertoire that allows selection into the B-1a population. As it is selected by self-antigen, the B-1 repertoire tends to be autoreactive. This potentially dangerous repertoire is also useful, as B-1 cells are essential for resistance to several pathogens and they play an important role in mucosal immunity. The CD5 molecule can function as a negative regulator of BCR signaling that may help prevent inappropriate activation of autoreactive B-1a cells.

INTRODUCTION Any one of three types of antigen—thymus independent type 1 (TI-1), thymus independent type 2 (TI-2), or thymus dependent (TD)—can initiate B cell proliferation, differentiation, and ultimately antibody secretion. B cells activated by antigens of different types enter distinct differentiation pathways. Thus, a B cell responding to a TD antigen plus CD40 ligand matures in a germinal center, is able to somatically mutate, and gives rise to memory cells as well as plasma cells. In contrast, a TI-1-responding cell produces IgM, perhaps IgG2b, no memory, and little or no somatic mutation. Thymus independent type 2 antigens induce multivalent cross-linking of the B cell receptor (BCR). This is not sufficient to drive responding B cells into antibody production because additional signals supplied by noncognate interaction 0732-0582/02/0407-0253$14.00

253

1 Feb 2002

10:15

Annu. Rev. Immunol. 2002.20:253-300. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

254

AR

AR152-10.tex

BERLAND

¥

AR152-10.SGM

LaTeX2e(2001/05/10)

P1: GJC

WORTIS

with T cells, NK cells, and perhaps other cells are necessary (1). Ultimately, TI-2responding B cells mainly produce IgM and some IgG3 antibodies containing few or no somatic mutations. They do not usually generate long-term memory cells (no secondary response) (2, 3). We propose that the phenotype of TI-2 responding cells differs from that of B cells triggered by a TD or innate (TI-1) antigen. We further propose that depending on the circumstances of interaction of the B cell with TI-2 antigen, the responding cells will be found in the spleen as marginal zone (MZ) B cells, in the peritoneal or pleural cavity as B-1a or B-1b cells, or as anergic cells destined for elimination. [Although differing in detail, a model broadly similar to ours was developed independently by Martin & Kearney and published earlier in an elegant review (4).]

Definition of B-1 Cells B-1 cells can be distinguished from all other B cells by surface phenotype. In contrast to recirculating follicular (also B-2 or B-0) cells, they are CD45 (B220lo), IgMhi, CD23−, CD43+, and IgDlo. They are also larger and exhibit more side scatter than do B-2 cells. In the peritoneal cavity (PerC), but not the spleen, B-1 cells express C3 (CD11b, Mac-1). B-1 cells are absent from peripheral lymph nodes (LN) and variably make up about 5% of splenic B cells. However, they constitute a substantial fraction of B cells in the peritoneal and pleural cavities (5, 6). Originally, B-1 cells were identified by their expression of CD5. Subsequently, a population of peritoneal CD5− B cells was identified whose surface phenotype was in other respects identical to that of B-1 cells. By consensus, CD5+ B-1 cells are referred to as B-1a cells and CD5− B-1 cells as B-1b cells. Most of the work reviewed in this article was with B-1a cells or did not distinguish between B-1a and B-1b. Furthermore, it is concerned almost exclusively with B-1a cells from the PerC. B-1a cells from the spleen have not been much studied because of their low frequency. However, they differ from PerC B-1a cells in two important respects. PerC B-1a cells contain constitutively active STAT-3, an inducible transcription factor (6a), while splenic B-1a cells do not (6b). Splenic B-1a cells, unlike those isolated from the PerC, flux calcium normally upon BCR ligation (26). The work reviewed here almost exclusively concerns mice. In humans there exist two classes of CD5+ B cells, only one of which appears to share other phenotypic properties with murine B-1a cells (6c). Like murine B-1a cells, CD5+ human peripheral blood B cells have been reported to produce polyspecific autoreactive antibody (173, 174). However they have not been much studied and it is not clear to what extent they are equivalent to B-1a cells in the mouse. In other species the situation is even more uncertain. CD5+ B cells have not been reported in rat (180) except for a recent report of rats expressing a transgenic neuropeptide Y (6d). In rabbits (6e), sheep (6f), cattle (6g), and chickens (6h), all or most peripheral B cells express CD5.

1 Feb 2002

10:15

AR

AR152-10.tex

AR152-10.SGM

LaTeX2e(2001/05/10)

P1: GJC

B-1 CELLS AND CD5

255

Properties of B-1 Cells

Annu. Rev. Immunol. 2002.20:253-300. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

In addition to surface phenotype, B-1 cells have a number of distinguishing properties discussed in more detail later in the review. In vitro, they are long lived (7), refractory to activation through B cell receptor (BCR) ligation (8–10), and, in contrast to B-2 cells, induced to proliferate by treatment with phorbol esters (11). The repertoire of B-1 cells is decidedly skewed toward reactivity with common bacterial and self-antigens. Typical B-1 immunoglobulin genes have fewer N insertions than those of most B-2 cells (12), and they usually do not contain somatic mutations (13).

Functions of B-1 Cells The segregation of B cells with a particular repertoire into a population with the B-1 phenotype is presumably of significance to the animal. Some aspects of the B-1 phenotype may assure that weakly autoreactive B cells are not recruited into germinal centers where affinity maturation could result in high-affinity, pathogenic autoreactivity. Other aspects of the phenotype may have been selected, as they enable responses to certain TI-2 antigens or provide natural antibody. We return to these issues after an examination of B-1 development.

ORIGINS OF B-1 CELLS B-1 and B-2 cells were originally proposed to derive from different, committed, precursors and therefore to represent the end products of two distinct lineages. This view was based on cell transfer studies in which fetal liver reconstituted both the B-1 and B-2 compartments of irradiated mice, while adult bone marrow was generally limited to the generation of B-2 cells (14–16). Subsequently, the fetal omentum (17) and paraaortic splanchnopleura (18) were shown to contain precursors exclusively for B-1 cells. Thus, the B-1 lineage appeared to be predominantly of fetal and the B-2 lineage of adult origin. That a fetal-derived B cell population would persist for the life of the animal was explained by additional studies showing that B-1 cells were self-renewing. Transfer of B-1 cells into neonatal (19) or irradiated (20) mice resulted in long-term reconstitution of the B-1 compartment (discussed in 14). According to this lineage model, the progenitors of B cells either have or do not have the potential to become a B-1 cell. We (21) and others (22, 23) have proposed an alternative model to account for the properties of B-1 cells. In this induced-differentiation model, the B-1 phenotype was proposed to be a consequence of TI-2-like activation. This model was based on our observation that BCR cross-linking, in the absence of an innate antigen, T cell help, or CD40 ligation, resulted in the induction of CD5 on splenic B-2 cells (21, 24). It was subsequently shown that BCR cross-linking also induced in B-2 cells the ability to proliferate in response to phorbol ester, another feature of B-1 cells (25). The combination of BCR ligation and IL-6 treatment induced two additional

1 Feb 2002

10:15

Annu. Rev. Immunol. 2002.20:253-300. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

256

AR

AR152-10.tex

BERLAND

¥

AR152-10.SGM

LaTeX2e(2001/05/10)

P1: GJC

WORTIS

features of B-1 cells, the downregulation of CD23 and of IgD (21). However, an in vitro treatment that induces the full B-1 phenotype has not been found, which suggests that additional unknown signals are involved. According to the induced differentiation model, encounters with naturally occurring TI-2 antigens account for the appearance of B-1 cells in vivo. Differences in the ability of precursors from adult and fetal sources to generate B-1 cells are due to differences in the repertoires and hence specificities of B cells derived from these sources. The fetal/neonatal repertoire is skewed toward the expression of immunoglobulins that bind frequently encountered TI-2 antigens. The adult repertoire rarely generates these specificities and therefore generates few B-1 cells. In contrast, the lineage model postulates that differences in the repertoires of B-1 and B-2 cells are a result of antigen-driven selection for survival or expansion. Interestingly, IgM expression per se is not required for B-1a development. Mudeficient mice that can still express IgD have a population of PerC CD5+ B cells that express high levels of surface IgD (25a). This is particularly striking given the fact that B-1 cells normally express only very low levels of IgD. The functional properties of these cells were not examined. We now review evidence, some of it very recent, that we believe is most consistent with the induced differentiation model of B-1 development.

B Cell Receptor Signaling is Essential for B-1 Development The study of gene targeted and transgenic mice has generated strong evidence that BCR signaling is critical for B-1 development. Mutations that disrupt BCR signaling result in substantial depletion of the B-1 subset while largely sparing B-2 cells (see Table 1). Conversely, mutations or transgenes that enhance BCR signaling result in an expanded B-1 compartment (Table 1). These results indicate that B-1 cells require a BCR-generated signal for development, survival, or expansion. If there is a B-1 lineage, the phenotype is not expressed unless there is ligation of the BCR by antigen.

B Cell Receptor Specificity Determines the Likelihood that a Given B Cell Acquires the B-1 Phenotype In mice transgenic for B-1-derived immunoglobulin genes, transgene expressing B cells are predominantly CD5+ (Table 2). In one recent study, such cells were shown to have the B-1 property of enhanced in vitro life span (26). Conversely, in mice carrying B-2-derived transgenes, transgene expressing B cells are almost exclusively of the B-2 phenotype (Table 2). The importance of BCR specificity in the generation of B-1 cells suggests that the requisite BCR signaling is ligand-dependent. Conclusive evidence that B-1 development is dependent on a ligand-driven process comes from recent work of Hayakawa, Hardy, and colleagues (27). They had previously cloned a B-1-derived hybridoma (SM6C10) expressing an anti-T cell antibody (ATA) specific for the T cell glycoprotein Thy-1 (28). When transgenic mice were made that expressed

1 Feb 2002

10:15

AR

AR152-10.tex

AR152-10.SGM

LaTeX2e(2001/05/10)

P1: GJC

B-1 CELLS AND CD5

257

Annu. Rev. Immunol. 2002.20:253-300. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

TABLE 1 Genetic alterations that affect B-1 cell numbers Alterations that decrease B-1 cell numbers Mutation of positive regulators of BCR signalingPoint mutation (xid) or deletion of btk Deletion of PKCβ Deletion of PLCγ Deletion of P85α of PI-3 Kinase Deletion of CD19 Deletion of BLNK (SLP-65) Deletion of CD21/35 Deletion of vav-1 Mutation of B cell transcription factors Deletion of Oct 2 Deletion of Aiolos Deletion of NFATc

(33, 251, 252) (253) (254, 255) (256, 257) (258, 259) (260–263) (264) (but see 265) (266, 267) (268) (269) (Berland & H. Wortis, unpublished observation)

Mutation of growth factors or growth factor receptors Deletion of IL-5 Deletion of IL-5R

(270) (271, 272)

Other Deletion of Cyclin D2

(273)

Alterations that increase B-1 cell numbers Mutation of negative regulators of BCR signaling Loss of function of SHP-1 (motheaten, motheatenv) Deletion of CD22 Deletion of Lyn Deletion of CD72

(274) (155, 156, but see 153, 154) (151) (275)

Overexpression of positive regulators of BCR signaling CD19 transgenic

(276)

Overexpression of transcription factors Fli-1 transgenic

(277)

Overexpression of growth factors IL-5 transgenic Osteopontin transgenic IL-9 transgenic

(278) (162) (279)

the SM6C10 heavy chain, they had high titers of ATA, all of which was produced by transgene-expressing B-1a cells (27). The majority of these cells expressed an endogenous light chain identical to that in the original SM6C10 hybridoma (27). Thus, as in other transgenic mice, expression of a B-1a immunoglobulin specificity correlated with acquisition of the B-1a phenotype. Strikingly, when SM6C10 µ transgenic mice were crossed onto a Thy-1−/− background, ATA producing B cells

1 Feb 2002

10:15

258

AR

AR152-10.tex

BERLAND

¥

AR152-10.SGM

LaTeX2e(2001/05/10)

P1: GJC

WORTIS

Annu. Rev. Immunol. 2002.20:253-300. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

TABLE 2 Effect of Ig-transgene expression on B cell development

Transgene

Specificity

Source

Phenotype of transgene expressing B cells

3-83

H-2KkDk

Normal Adult Spleen

B-2

(26, 280)

HyHEL-10

Hen Egg Lysozyme

Normal Adult Spleen

B-2

(40, 281)

References

VH81X

Unknown

Normal Fetal Liver

MZ

(109)

M167

Phosphoryl choline

Normal Adult Spleen

MZ

(109)

VH12/Vκ4

Phosphatidyl choline

CD5+ Lymphoma

B-1a

(32)

VH11/Vκ9

Phosphatidyl choline

B-1 cell

B-1a

(26)

SM6C10

thy-1

B-1 cell

B-1a

(27)

4C8

Mouse Red Blood Cells

NZB Spleen

B-1

(119)

2-12H

sM snRNP

MRL/lpr mouse

B-1a

(37)

and transgene expressing B-1 cells were absent (27, 29, 30). Thus, engagement of the BCR by antigen, in this case Thy-1, is necessary for B-1a development. Ligand-mediated BCR signaling could be required to induce the B-1 phenotype or for the survival or expansion of cells already committed to the B-1 lineage. Evidence for the former comes from a series of studies in Steve Clarke’s laboratory. This group has been working with mice carrying B-1-derived rearranged genes encoding the heavy (VH12) and light (Vκ4) chains, which together produce an antibody specific for phosphatidyl choline (PtC). Antibody to PtC is frequently produced by B-1 but not B-2 cells (31). In VH12/Vκ 4 transgenic mice, B cells expressing both transgenes are almost exclusively B-1 (32). In order to determine if B-1 commitment preceded or followed transgene expression, Clarke & Arnold bred the X-linked immune deficiency (xid) mutation into these mice. Xid is a mutation in Bruton’s tyrosine kinase (Btk), a gene which encodes a kinase necessary for BCR signaling. Xid is known to prevent B-1, but not B-2, development (33, 34). As expected, in transgenic xid mice, there were no transgene-expressing B-1 cells (23). Significantly, transgene-expressing B-2 cells (B-0, in the nomenclature of Clarke and colleagues) were now found (23). The most straightforward interpretation of this result is that on a wild-type background, transgene expression in a B-2 (B-0) cell rapidly results in the acquisition of the B-1 phenotype. On an xid background, B-1 differentiation cannot occur because there is no effective BCR signaling, and this allows accumulation of transgene-expressing B-2 (B-0) cells. However, in xid mice most B cells are transitional or immature, not mature. The argument can then be made that this developmental arrest prevents lineage committed B cells from maturing into B-1 cells (35). In a subsequent study, Arnold et al. identified splenic B cells in VH12/Vκ 4 transgenic mice that had a phenotype intermediate between B-1 and B-2 (B-1int) (36).

1 Feb 2002

10:15

AR

AR152-10.tex

AR152-10.SGM

LaTeX2e(2001/05/10)

P1: GJC

Annu. Rev. Immunol. 2002.20:253-300. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

B-1 CELLS AND CD5

259

B-1 cells are CD5+, CD43+, IgMhi, FSChi, CD45(B220)lo, and CD23−, where low and high refer to staining intensity relative to B-2 cells. B-1int cells are CD5+, CD23int, CD45(B220)int, and FSCint. B-1int cells also have IgM and CD43 levels slightly lower than in B-1 cells. When these intermediate phenoptype cells were transferred into sublethally irradiated mice and analyzed five days later, they had acquired the B-1 phenotype, with the exception of the loss of CD23 expression. This differentiation to B-1 was blocked by cyclosporin A (CSA) treatment, which suggests that it requires BCR signaling (see below) (36). The discovery of B-1int cells that can progress to a fuller B-1 phenotype in a CSA sensitive manner is consistent with the notion that B-1 cells are derived from cells with a B-2 (or B-0) phenotype after BCR ligation. Presumably B-1int cells are detectable in VH12/Vκ4 transgenic mice, but not in nontransgenic mice, because of the much larger number of B cells undergoing differentiation to B-1 in the transgenic mice. Using a different immunoglobulin transgenic model, Clarke’s group has recently shown that splenic B cells with a B-2 phenotype can differentiate into PerC B-1 cells. These mice express an anti-Sm snRNP heavy chain derived from an MRL/lpr mouse. Transgenic, anti-Sm B cells are found in the spleens as transitional, rapidly turning over B-2 cells and in the PerC as B-1a cells. Splenic B-2 cells assume a B-1a phenotype after transfer to sublethally irradiated recipients (37). One important question is the nature of the signals that drive B-1 differentiation. The nature of the B-1 repertoire, enriched as it is for weak autoreactivity, suggests that B-1 cells are usually selected by weak interactions with self-antigens. This is further supported by the example of anti-Thy-1 transgenic mice discussed above. However, when mice expressing transgenes encoding an antibody (3-83) with specificity for H-2Kk were crossed with mice expressing ligands with very low affinity for the 3-83 immunoglobulin (KA ∼ 1 × 104 M−1), the progeny deleted or receptor edited the 3-83+ B cells (38). Thus, as suggested by Chumley et al. (26) either B-1 cells are selected by even weaker autoreactivities, or the signals driving B-1 cell development are of a qualitatively different nature than those leading to tolerance. One possibility is that cells targeted for tolerance by self-reactivity can be rescued by an additional signal or signals that drive them into the B-1 compartment. This is suggested by work from Tim Behren’s laboratory examining tolerance in anti-HEL/sHEL transgenic mice (39). In this system, mice express both an immunoglobulin specific for hen egg lysozyme (HEL) and a gene encoding a soluble form of HEL (sHEL). Anti-HELexpressing B cells become anergic as a result of encountering sHEL (40). Behrens and colleagues recently showed that these anergic cells express CD5, albeit at a level below that seen in typical peritoneal B-1 cells (39). Also, unlike B-1 cells, these anergic cells express low levels of surface IgM (39). Whatever the precise nature of the signals leading to the B-1 phenotype, B-1 development appears rather sensitive to the strength of the signal received from the BCR. This is suggested by two studies showing that in mice transgenic for a B-1 BCR specificity, reduction of BCR surface density, even by a factor of only two, significantly reduces the number of transgene expressing B-1 cells (41, 42).

1 Feb 2002

10:15

260

AR

AR152-10.tex

BERLAND

¥

AR152-10.SGM

LaTeX2e(2001/05/10)

P1: GJC

WORTIS

Annu. Rev. Immunol. 2002.20:253-300. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

Adult Bone Marrow Can Give Rise to B-1 Cells Cell transfer studies clearly indicate that fetal precursors more readily give rise to B-1 cells than do adult precursors. However, examples exist in which B-1 cells were generated after adoptive transfer of adult bone marrow cells. The Herzenberg laboratory reported sporadic generation of PerC B-1 cells after transfer of adult bone marrow to SCID recipients. They attributed this to the possible persistence of fetal precursors in some preparations of adult bone marrow (43). We reported recovery of splenic B-1 cells after transfer of adult (>5 month old) bone marrow into neonatal SCID recipients (44). In this study, reconstituted B-1 cells were shown to have a high frequency of N-insertions, suggesting that they truly were derived from adult precursors (see below). Others, too, have reported that adult bone marrow contains some progenitors for B-1 cells (45–47). Finally, Whitmore et al. reported that transferred bone marrow (but not PerC cells) gave rise to cells that responded to the TI-2 antigen polyvinyl pyrrolidinone (PVP) after immunization, and that the anti-PVP antibody-producing cells had acquired the B-1 phenotype (48). These workers also demonstrated the (likely) presence of N-insertions in 10 of 16 hybridomas derived from these anti-PVP producing cells. Despite these results, the ability of adult precursors to give rise to B-1 cells remained in question (see for example 49). These doubts were in part based on studies showing that purified BM pro-B cells that had begun but not yet completed heavy chain rearrangement could give rise to B-2 but not B-1 cells after transfer into SCID recipients (50). Conversely, pro-B cells at the same developmental stage isolated from day-16 fetal liver gave rise predominantly to B-1 cells. Similar results were obtained when pro-B cells were allowed to differentiate in vitro in stromal cell cultures (50). However, a recent study indicates that the ability to generate B-1 cells is not uniquely a property of fetal precursors. In this study, mice were made transgenic for VH11/Vκ 9, a B-1-derived anti-PtC specificity. As in other studies (Table 2), expression of these B-1-derived transgenes resulted in B cells with the B-1 phenotype (26). Significantly, CD45(B220)−/CD19− precursors isolated from the bone marrow of adult VH11/Vκ 9 transgenic mice exclusively gave rise to B-1 cells when transferred into irradiated recipients or placed in culture with IL-7. These cells had the surface phenotype and enhanced ability to survive in culture typical of B-1 cells. In contrast, equivalent cells isolated from the bone marrow of mice transgenic for 3-83, a B-2 specificity, gave rise exclusively to B-2 cells, both in vivo and in vitro. This study would seem to demonstrate beyond any doubt that adult BM precursors, when expressing an appropriate specificity, differentiate into CD5+ long-lived cells, i.e., B-1 cells.

Fetal vs. Adult B Cell Development The fact remains that fetal precursors more readily give rise to B-1 cells than do adult precursors. Based on the studies discussed above, this most likely reflects differences in repertoire between fetal and adult-derived B cells. That such differences exist is well established. Pro- and pre-B cells in the fetal liver rearrange

1 Feb 2002

10:15

AR

AR152-10.tex

AR152-10.SGM

LaTeX2e(2001/05/10)

P1: GJC

Annu. Rev. Immunol. 2002.20:253-300. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

B-1 CELLS AND CD5

261

their heavy chains with the preferential use of J-proximal V genes (51). V gene usage in adults is approximately random, that is, the frequency of use of genes from a given VH family is proportional to the number of VH genes in the family (52). In adults, terminal deoxynucleotide transferase (TdT) makes nontemplated nucleotide insertions (N insertions) during VDJ rearrangements. This enzyme is not expressed in fetal pre-B cells (53, 54). As a result, fetal-derived B cells contain germ-line encoded heavy and light chains, whereas adult-derived B cells contain heavy and light chains that diverge from germ line. Recent evidence indicates that fetal vs. adult differences in V region usage are selected for during pre-B cell development (55). This was seen by examining productive and nonproductive rearrangements of heavy chain genes in single, sorted pre-B cells from adult BM. In nonproductively rearranged alleles there is, as in the expressed repertoire of neonatal B cells, preferential use of J-proximal V genes. This skewing is absent in productively rearranged alleles, indicating that there is either selection against use of J-proximal V regions or selection for use of other V regions. This selection is dependent on surrogate light chain (SLC) expression since in pre-B cells of λ5 knockout mice, which lack a functional SLC, both productive and nonproductive rearrangements exhibit preferential use of J-proximal V genes (55). The SLC, which consists of V-pre B and λ5, associates with µ heavy chains to form the pre-BCR. Successful assembly of this complex is necessary for light chain rearrangements to begin (56; but see below for a possible exception to this). Evidence to support the role of the SLC in randomizing V gene usage comes from examination of the J-most heavy chain gene, VH81X. This gene is common in the neonatal repertoire but rare in the adult (57, 58). In adults productive rearrangements of this gene were seen only in pre-B cells that expressed c-kit, the receptor for stem cell factor (55). C-kit is only found on early B cell progenitors, at a stage prior to SLC-dependent pre-B cell expansion (59). Of seven productively rearranged VH81X heavy chains cloned from adult pre-B cells, none was able to associate with the SLC when transfected into an SLC expressing cell line (55). What then is the basis for the successful generation of VH81X expressing B cells in the neonate? One possibility is that the germ line–encoded VH81X gene can associate with the SLC, but that this is prevented by introduction of N insertions. This notion is supported by the fact that an N-less, fetal-derived VH81X transgene could efficiently drive adult B cell development (60). In addition, introduction of this VH81X transgene into λ5 knockout mice did not relieve the block to B cell development caused by the absence of λ5 (60). Thus, this transgene cannot promote B cell development in a λ5-independent manner. Additional evidence that it is N insertion that prevents VH81X expression in adult B cells comes from transgenic mice in which expression of a TdT transgene is forced in fetal pre-B cells. In nontransgenic neonatal mice the ratio of productive to nonproductive VH81X rearrangements (P/NP) was 2.4, indicating strong selection for this gene. In TdT transgenic neonates, the P/NP dropped to 0.32, indicating that N insertion inhibits selection for VH81X heavy chains (61). Taken together,

1 Feb 2002

10:15

Annu. Rev. Immunol. 2002.20:253-300. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

262

AR

AR152-10.tex

BERLAND

¥

AR152-10.SGM

LaTeX2e(2001/05/10)

P1: GJC

WORTIS

these studies suggest that in the course of evolution the VH81X transgene was selected to be maintained in its germ-line configuration. This may be generally true of members of the fetal/neonatal repertoire. For these genes, the absence of N insertions may be sufficient to explain their restriction to fetal/neonatal B cells. Recently, Wasserman et al. proposed another possible mechanism to ensure the generation of different repertoires in fetal/neonatal and adult B cells (62). They found that a B-1-derived VH11 heavy chain is unable to associate with the SLC, yet contributed to the fetal-derived mature B cell population. They went on to suggest that it is the inability of this heavy chain to form a pre-BCR that promotes expansion of fetal pre-B cells (62). In contrast, B-2-derived heavy chains were able to form a pre-BCR and inhibited the expansion of fetal pre-B cells (62). This contrasts with the situation in adult bone marrow where association with SLC has been shown to be necessary for B cell development. In other words, Wasserman et al. propose, pre-B cells expressing a heavy chain/SLC complex fail to expand in a fetal microenvironment, but the complex is required for expansion in the adult marrow. These conclusions are hard to reconcile with the demonstration that in fetal liver organ cultures, efficient B cell development is dependent on λ5 expression (63) or with the observation that a VH81X transgene failed to drive pre-B cell development in λ5 knockout mice (60). In addition, Ye et al. demonstrated that a B-1-derived VH12 heavy chain was able to associate with the SLC and form a pre-BCR (64). Despite the superior ability of fetal precursors to generate B-1 cells, the B-1 repertoire in adult mice is not the same as the fetal/neonatal B cell repertoire. For example, VH81X, commonly expressed on fetal/neonatal B cells, is largely absent from the adult. This may reflect the loss in the adult of a self-antigen necessary for the maintenance of VH81X-bearing B-1 cells because after transfer into adult SCID mice, PerC B-1 cells transgenic for VH81X failed to expand and survive (60). Furthermore, analysis of VDJ genes isolated from PerC B-1 and B-2 cells by single cell PCR found that about 60% of genes cloned from B-1 cells contained N-insertions (12). Although this contrasts with genes cloned from B-2 cells, where greater than 90% contain N insertions (12), it suggests that many B-1 cells may be derived from adult precursors. Preferential use of J-proximal V genes was not observed in B-1 cells (12), again a fact consistent with the possible adult origin of many B-1 cells, although this may also be explained by post Ig–rearrangement selection. Any contribution to the B-1 compartment by newly arising adult B cells is constrained by the existence of negative feedback regulation of B-1 development (65).

Other Factors Influencing B-1 Development While appropriate antigen receptor signaling is crucial for the generation of B-1 cells, other signals also play important roles in their development and localization. B-1 cells, unlike B-2 cells, constitutively express the IL-5 receptor (65a). IL-5 transgenic mice that overexpress IL-5 have increased numbers of B-1 cells (278). Conversely, mice containing targeted disruptions of the IL-5 (270) or IL5R (271)

1 Feb 2002

10:15

AR

AR152-10.tex

AR152-10.SGM

LaTeX2e(2001/05/10)

P1: GJC

Annu. Rev. Immunol. 2002.20:253-300. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

B-1 CELLS AND CD5

263

genes have decreased numbers of B-1 cells early in life. B-1 numbers recover by 6–8 weeks of age, indicating that another factor can substitute for IL-5 later after birth. IL-9 might be the relevant factor, as its overexpression causes an increase in the fraction of B-1 cells (279). A subfamily of TNF-related ligands and their receptors have been discovered to be important in B cell development and function. These ligands are BLys (also called TALL-1, BAFF, THANK, and zTNF4) and APRIL. They both bind with high affinity to the receptors TACI and BCMA (reviewed in 65b). Disruption or overexpression of these ligands and receptors results in complex B cell phenotypes that are discussed in Siegel and Lenardo (65c). BlyS, is essential for B-2 but not B-1 development (65d). Nonetheless, in one study, BlyS overexpression resulted in an expanded splenic B-1 population (159). Perhaps because they express higher levels of the CXCL13 receptor CXCR5 than do splenic B-2 cells, PerC B-1 cells are more sensitive to this chemokine (65e). In mice containing a targeted disruption of the CXCL13 gene there was defective homing of B-1 cells to the PerC (K. Ansel and J. Cyster, personal communication). Splenic B-1 cells were still present. The responses of PerC, but not splenic, lymphocytes (B-1, B-2, and T), to CXCL13 (as well as to another chemokine, CCL21) are reduced in aly/aly mice which contain a point mutation in the NIK tyrosine kinase gene (85). This point mutation results in defective processing of NF-κB2 (65f ).

FUNCTIONS OF B-1 CELLS Production of Natural Serum IgM B-1 cells adoptively transferred into irradiated mice (43, 66), unmanipulated neonatal mice (19), or B cell–depleted neonatal mice (65) produce quantities of IgM approximating the levels seen in unmanipulated animals. Based on such studies, B-1 cells are believed to be the primary source of natural IgM. This antibody, which is produced in the absence of exogenous antigenic stimulation (67, 68), is polyreactive, weakly autoreactive, and reactive with many common pathogen-associated carbohydrate antigens. Several recent reviews highlight the important roles of natural antibody in adaptive immune responses, protection from bacterial infection, and protection from autoimmunity as well as in a model of ischemia reperfusion injury (69, 70). Consistent with a major role for B-1 cells in natural IgM production, a number of natural IgM specificities have been identified in the B-1 repertoire. These include specificities for phosphorylcholine (PC) (71), phosphatidyl choline (PtC) (31, 72, 73), thymocytes (28), LPS (74), and influenza virus (75). In mice mutated so that they express membrane IgM (mIgM) but are unable to produce secretory IgM (sIgM), there is an increase in the number of B-1 cells, suggesting that the size of the B-1 compartment is regulated by serum IgM levels (76, 77). The ability to produce natural IgM is not a property unique to cells with the B-1 phenotype. Purified small resting lymph node B cells have been shown to

1 Feb 2002

10:15

Annu. Rev. Immunol. 2002.20:253-300. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

264

AR

AR152-10.tex

BERLAND

¥

AR152-10.SGM

LaTeX2e(2001/05/10)

P1: GJC

WORTIS

differentiate into long-lived activated IgM-producing cells and to reconstitute normal serum IgM levels after transfer into SCID recipients (78). These cells remained CD5− over the course of the experiment. Interestingly, they acquired a phenotype similar (and perhaps identical) to that of MZ B cells (78). Once established, these IgM producing, B-2-derived cells were resistant to replacement by a second injection of purified B-2 cells. They also resisted replacement by newly emerging B cells when adult bone marrow was injected along with LN B cells. In these mice, a normal-size B cell compartment, derived from injected bone marrow, coexisted with an activated, IgM-secreting population largely derived from injected lymph node B cells (78). Replacement did slowly occur, possibly reflecting competition between established and newly emerging cells for access to factors needed for differentiation into IgM-secreting cells. These experiments suggest that it may be the repertoire of B-1 cells, rather than the B-1 phenotype per se, that selects them into the pool of natural IgMsecreting cells. As discussed earlier, this repertoire also drives cells to assume the B-1 phenotype. It would be interesting to see if injected B-1 cells would be more efficient than mature B-2 cells in displacing lymph node–derived IgM-producing cells in experiments similar to those discussed above. Several studies demonstrate the importance of demonstrably B-1-derived natural antibody. For instance, antibodies expressing the T15 idiotype, which are almost exclusively of B-1 origin (71), play an important role in protection from infection by Streptococcus pneumoniae (79, 80). Furthermore, it was recently shown that antibodies with the T15 idiotype bind oxidized low-density lipoprotein (LDL) (81). This binding may aid in clearance of LDL from the blood and thus play a role in prevention of atherosclerosis. Alternatively, these antibodies may bind to atherosclerotic plaques, initiate an inflammatory response, and contribute to the pathogenesis of vascular disease. Anti-PtC antibody is critical for protection from acute septic peritonitis after cecal ligation and puncture (CLP) (82). Mice unable to secrete IgM (sIgM−/−) because of a targeted mutation are highly susceptible to death after CLP. Transfer of normal mouse serum or purified B-1-derived monoclonal anti-PtC protected sIgM−/− mice from CLP-induced death (82). B-1 cells are responsible for almost all natural antibody reactive with LPS (74). This antibody is important in clearance of LPS as indicated by the increased susceptibility of antibody-deficient mice to death after LPS injection (83). Administration of normal mouse serum prior to LPS injection protects mice from death (83). Uninfected mice contain natural anti-influenza IgM, which is derived exclusively from B-1 cells (75). Upon influenza infection, titres of this antibody do not increase. Rather, an increase in B-2-derived immune IgM is observed (75). Reconstitution experiments demonstrate that resistance to influenza infection is dependent on both B-1- and B-2-derived IgM (84). In the absence of B-1-derived natural anti-influenza IgM there is a delayed T-dependent IgG2a response and increased mortality. Thus, natural IgM is necessary for a normal adaptive response to influenza (84). Similarly, IgG responses to model TD antigens are impaired in

1 Feb 2002

10:15

AR

AR152-10.tex

AR152-10.SGM

LaTeX2e(2001/05/10)

P1: GJC

B-1 CELLS AND CD5

265

sIgM−/− mice (76, 77) but can be rescued by administration of normal serum IgM prior to immunization (77). The role of natural IgM in adaptive responses could be to facilitate complementmediated localization of antigen to germinal center follicular dendritic cells (FDCs), to lower the threshold of B cell activation by allowing simultaneous engagement of the BCR and the CD21/CD19 complement receptor, or both (see 69 for discussion).

Annu. Rev. Immunol. 2002.20:253-300. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

Production of IgA in the Mucosal Immune System Several groups demonstrated that transfer of total peritoneal B cells (66, 85), or sorted B-1 cells (86), into lethally irradiated (66, 86), or RAG2−/− (85) mice resulted in the appearance of donor-derived IgA plasma cells in the gut lamina propria (LP) and mesenteric lymph nodes (MLN). Transfer of fetal omentum into SCID mice, which reconstituted the B-1 but not the B-2 compartment, also resulted in the appearance of IgA secreting plasma cells in the LP (17). In mixed radiation chimeras, where the relative contributions of PerC- and BM-derived donor cells could be followed, PerC cells were estimated to give rise to about 40% of the IgAproducing cells in the gut (66). In a second, independent study in which purified B-1 cells were transferred, a similar fraction of LP IgA producing cells appeared to be B-1 derived (86). The role of B-1 cells in gut IgA production is further supported by the fact that CD19-deficient and xid mice, both of which have a selective reduction in B-1 cell numbers, have decreased numbers of IgA-secreting plasma cells in the LP (but not in the spleen) (86). IgA production is dependent on exposure to exogenous stimuli because mice raised under germ-free conditions largely lack gut IgA (67, 86–88). This is in contrast to B-1-derived natural IgM, which is produced even in germ-free mice (67, 87, 88). B-1, but not B-2, gut IgA production is independent of T cell help because it is unaffected by mutations that eliminate T cells (nu/nu or TCRβ −/−/ δ −/−) (86). Whether T cell–independent IgA production by B-1 cells requires particular gut associated lymphoid structures was addressed by Macpherson et al., who examined TNFR-1−/− mice (86). TNFR-1−/− mice have lymph nodes (LNs), but their LNs lack FDCs, primary B cell follicles, and GCs (89). They either lack PPs or contain a reduced number with altered architecture (89, 90). Despite these defects, TNFR-1−/− mice have near normal numbers of LP IgA secreting cells (86), which suggests that neither B-1- nor B-2-derived IgA production requires organized follicular lymphoid structures. This raises the question of whether LNs or PPs are required at all for gut-associated IgA. However, mice lacking all LNs and PPs due to targeted disruption of the LTα gene (91, 92), or double knockout of LTα and TNF genes (93), exhibit severely reduced gut IgA production (92, 93). Thus, both B-1 and B-2 IgA production is compromised by the loss of LNs and PPs, despite the fact that B-1 IgA production is independent of T cells. LTβ −/− mice (94), or mice treated in utero with an LTβ-Ig fusion protein (95, 96), lack PPs and peripheral LNs, but they do contain MLNs. In a recent

1 Feb 2002

10:15

Annu. Rev. Immunol. 2002.20:253-300. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

266

AR

AR152-10.tex

BERLAND

¥

AR152-10.SGM

LaTeX2e(2001/05/10)

P1: GJC

WORTIS

study, gut IgA production in LTβ-Ig treated mice (PP−/MLN+) was compared to that of TNF−/−LTα −/− mice (PP−/MLN−) (97). TNF−/−LTα −/− mice were again found to produce very little gut IgA. In contrast, LTβ-Ig-treated mice had both a nearly normal frequency of gut-associated IgA-producing cells and nearly normal gut IgA production. This suggests that MLNs provide something necessary for gut IgA production, including IgA production from B-1-derived cells. However, these data conflict with those of Koni et al. who found that LTβ −/− mice (also PP-/MLN+) produce very little fecal IgA (94). Perhaps LTβ has a function in gut IgA production independent of its role in MLN formation, and LTβ-Ig treatment does not interfere with this putative function Another strain of mouse with defects in lymphoid organ development is the naturally occurring alymphoplasia (aly) mouse. These mice carry a point mutation in the tyrosine kinase NIK (98) that results in the absence of all LNs and PPs as well immune deficiencies (99), including the absence of IgA-secreting plasma cells in the lamina propria and MLN (85). These mice also have increased numbers of PerC B-1 cells (85). PerC cells from aly/aly mice, in contrast to those from aly/+ mice, failed to generate IgA-producing cells in the lamina propria or MLN after transfer to Rag2−/− mice (85). Thus, defects in gut IgA production are at least partly a PerC cell-intrinsic effect. Recipients of aly/aly cells also had the elevated PerC B-1 cell levels characteristic of aly mice (85). This suggested the possibility that failure of B-1 cells to emigrate from the PerC might at least partly explain the lack of gut IgA-producing cells in aly mice. This idea is supported by the fact that aly/aly peritoneal B cells (both B-1 and B-2) have defective in vitro chemotactic responses to two chemokines; secondary lymphoid tissue cytokine (CCL21) and B lymphocyte chemoattractant (CXCL13) (85). Reduced chemotaxis was not seen in splenic B-2 cells from aly/aly mice. Consistent with an important role for CXCL13 in B-1 cell homing, it was recently reported that in wild-type mice PerC B-1 cells exhibit an enhanced migration response to CXCL13 in vitro and express higher levels of CXCR5, the receptor for CXCL13, compared to splenic B cells (65e). What is the significance of B-1-derived sIgA in the gut? Given the difference in repertoire of peritoneal B-1cells versus splenic B-2 cells, it is natural to ask if IgA derived from B-1 cells has a different range of specificities than that derived from B-2 cells. Support for this possibility comes from a study of mice containing allotype marked B-1 and B-2 cells (100). Examination of fecal bacteria from these mice using flow cytometry revealed that about 65% of bacteria were coated exclusively with B-1-derived IgA. Thirty per cent of bacteria were coated exclusively with B-2-derived IgA, and only about 5% were coated with IgA from both sources (100). It is unclear how this situation arises, or what its significance is. Bos et al. suggest that B-1-derived sIgA helps maintain commensal bacteria in the gut, whereas B-2-derived sIgA, because of recognition of different epitopes and/or binding with higher affinities, mediates elimination of potentially pathogenic bacteria (100).

1 Feb 2002

10:15

AR

AR152-10.tex

AR152-10.SGM

LaTeX2e(2001/05/10)

P1: GJC

B-1 CELLS AND CD5

267

Annu. Rev. Immunol. 2002.20:253-300. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

Role in TI-2 Responses Several lines of evidence suggest that B-1 cells are involved in TI-2 responses: (a) Their repertoire is skewed toward reactivity with TI-2 antigens; (b) they lack the somatic mutations that are the hallmark of affinity maturation in germinal centers; (c) xid mice, which lack B-1 cells, are unable to respond to TI-2 antigens; (d ) they produce IgM and IgG3, the major isotypes produced in response to TI-2 antigens; (e) TI-2, but neither TD nor TI-1, activation of B cells in vitro results in a partial induction of the B-1 phenotype. B-1 cells have in fact been shown to participate in TI-2 responses. Transferred B-1 cells respond in vivo to the TI-2 antigen α1-3 dextran (19). The best established role for B-1 cells in a TI-2 response is the production of T15 idiotype anti-PC antibody in response to infection by S. pneunoniae (71, 103, 110). This occurs even in TCRα −/−δ −/− mice, proving its T independence (103). The production of T15 antibody generated in response to S. pneumonia infection is in addition to the constitutive T15 antibody B-1 cells produce as a component of natural antibody. The T15 idiotype is generated by the association of unmutated VH1 and Vκ22 heavy and light chains. The fact that the murine germ line encodes such a useful specificity, that B cells expressing this specificity are selected and maintained in the apparent absence of external antigen, and that these B cells are readily recruited to respond to antigen in the absence of T cell help, has led to the proposal that B-1 cells are in essence “natural memory” cells (4). B-1 cells are not the only cells able to respond to TI-2 antigens. In fact, B-1 cells do not participate in the response to immunization with the TI-2 antigens TNP-Ficoll (72) and NP-Ficoll (19). Rather, splenic marginal zone (MZ) B cells appear to respond to these TI-2 antigens (104). Consistent with a role for MZ B cells in some TI-2 responses, Pyk-2−/− mice, which have normal numbers of PerC B-1 cells and MZ macrophages but lack MZ B cells, exhibit severely impaired responses to both TNP-Ficoll and dextran (107). MZ B cells are another phenotypically and functionally distinct subset of B cells (reviewed in 4). Entry into the MZ subset also appears to be a consequence of antigen receptor specificity. Thus, in M167 and VH81x µ transgenic mice, B cells expressing, respectively, the M167 and 35-1 idiotypes are greatly enriched in the MZ subset and largely missing from the recirculating follicular population. In contrast, in MD2 µ transgenic mice, B cells expressing an anti-HEL specific BCR are under-represented in the MZ subset (109). MZ B cells have been shown to be activated more rapidly by polyclonal stimuli than are follicular B cells, both in vitro (108) and in vivo (103). This suggests that this subset is comprised of cells selected based on their antigen-specificities and poised to provide a rapid first line of response to pathogens. Together with B-1 cells, they provide “natural memory” (4). What are the respective roles of B-1 and MZ B cells in TI-2 responses? A recent study by Martin et al. casts light on this question (103). They examined the response of M167 idiotype bearing B cells to infection by S. pneumoniae in M167 µ transgenic mice. M167 idiotype antibodies react with PC on S. pneumoniae. Using

1 Feb 2002

10:15

Annu. Rev. Immunol. 2002.20:253-300. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

268

AR

AR152-10.tex

BERLAND

¥

AR152-10.SGM

LaTeX2e(2001/05/10)

P1: GJC

WORTIS

adoptive transfer of sorted cells, Martin et al. demonstrated that the M167 id response to S. pneumoniae was almost exclusively due to the rapid differentiation of MZ B cells into IgM plasmablasts in this transgenic system. This is perhaps not surprising since prior to immunization 83% of M167 id+ cells in these mice had a MZ phenotype (109). However, when equal numbers of MZ and FO M167 id+ cells were transferred, MZ cells were more efficient at generating M167 id+ plasmablasts than were FO cells (103). Despite the ability of M167 id+ MZ B cells to respond to immunization with S. pneumoniae, most antibody to these organisms is produced by B-1 derived plasmablasts secreting T15 id+ IgM. To better understand what determines the magnitude of the B-1 vs MZ B cell response, Martin et al. transferred increasing numbers of M167 id+ MZ cells along with excess non-transgenic splenocytes. When small numbers of transgenic id+ MZ cells were transferred, the predominant response to iv immunization with S. pneumoniae was the secretion of T15 antibody. With increasing numbers of transferred transgenic MZ cells, there was an increase in the M167 response and a concomitant decrease in the T15 response. This suggests that the relative number of cells in MZ and B-1 subsets able to respond to an antigen determines, at least in part, the relative contribution of each subset to immunization with that antigen. In the experiment just discussed, antigen was administered iv and the response of splenic cells examined. T15 id+ B-1 cells exist in the spleens of unimmunized mice and presumably are the source of the T15 producing plasmablasts observed after iv immunization. However, since B-1 cells are enriched in the PerC, Martin et al. went on to ask whether this anatomical segregation of B-1 cells had any functional consequences. They found that although iv administration of S. pneumoniae led to production of both M167 and T15 antibodies, low dose ip administration resulted in a T15 response only in the PerC. It failed to elicit any splenic response, presumably because antigen was cleared by the PerC response and never became blood borne. In contrast, high dose ip administration of S. pneumoniae gave rise both to a T15 PerC response and a T15 plus M167 response in the spleen. Thus B-1 cells, because of their enrichment in the PerC, are ideally suited to respond to antigens that enter the organism through the gut epithelium. In contrast, MZ B cells are situated so as to be able to efficiently encounter and respond to blood borne pathogens. To the extent that different antigens may be encountered on blood borne versus gastrointestinal pathogens, it would be useful to select different repertoires into the MZ versus B-1 subset. It is unclear to what extent this is the case since little is known about the MZ repertoire. It is however true that some particular heavy/light chain combinations, such as that responsible for generating the T15 idiotype, are found only on B-1 cells (71). Since MZ and B-1 cells are selected into their respective compartments based on BCR specificity, it would be interesting to know what signals are responsible for this selection and how they differ between the subsets. In the case of the response to S. pneumoniae discussed above, both T15 B-1 cells and M167 MZ B cells are responding, as far as is known, to the same antigen on the bacterium, PC. Since

1 Feb 2002

10:15

AR

AR152-10.tex

AR152-10.SGM

LaTeX2e(2001/05/10)

P1: GJC

Annu. Rev. Immunol. 2002.20:253-300. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

B-1 CELLS AND CD5

269

PC is also present on mouse cells, it is plausible, but by no means certain, that PCreactive B-1 and MZ B cells are initially selected into their respective compartments by interaction with this self-antigen. This raises the question of why expression of each idiotype results in selection into a different compartment. This could be due to different affinities of these antibodies for PC, with T15 antibodies having a higher affinity for PC than do M167 antibodies (105). Selection into the B-1 compartment might require higher strength BCR signaling than selection into the MZ (or FO) population. Alternatively, selection of a cell into the MZ or B-1 compartment may depend on other, unknown signals in addition to BCR ligation, and the required signals may be different for B-1 and MZ differentiation. Recently Cariappa et al. reported results that they interpreted as indicating that B cells assume the MZ phenotype as a consequence of receiving lower levels of BCR signaling than are required for entry into the recirculating FO B cell pool (106). This would appear to contradict the idea that MZ B cells, in contrast to FO B cells, are a (self-) antigen selected population. Their conclusions are based on the fact that mice lacking the transcription factor aiolos have B cells that are hyperresponsive to BCR signaling and essentially lack MZ B cells. In contrast, xid mice, which have defective BCR signaling, have normal numbers of MZ B cells. (The authors show that there is an increase in the number of putative MZ precursors in xid mice; however, the fraction of cells with the full MZ phenotype is essentially the same as in WT.) Other interpretations of their data are possible. Aiolos deficient mice, despite the hyperresponsive phenotype of their B cells, have decreased numbers of B-1 cells (269). Yet, as discussed above, there is compelling data showing that B-1 development is dependent on stronger, rather than weaker BCR signaling. The effect of aiolos deficiency on positive selection of B-1 and MZ B cells may be independent of its effect on sensitivity of BCR signaling in vitro. As for xid mice, Martin and Kearney have shown that on an xid background VH81x transgenic B cells expressing the 35-1 id fail to be selected into the MZ subset in contrast to such cells on a WT background (109). This suggests that although xid mice have a normal sized MZ compartment, the repertoire of cells in this compartment may be altered. In conjunction with the absence of B-1 cells in xid mice, this could explain the inability of xid mice to respond to TI-2 antigens. Also inconsistent with weaker BCR signaling being necessary for MZ B cell development is the fact that both MZ and B-1 cells are absent from CD19 knockout mice (109, 259, 276). Distinct from the question of whether or not B-1 cells respond to TI-2 antigens is the issue of whether or not the response to TI-2 antigens induces a B-2 cell to express the B-1 phenotype as suggested by in vitro experiments. In an attempt to address this issue, Houghton and colleagues reconstituted lethally irradiated mice with allotype marked B-1 and B-2 cells, immunized them with the TI-2 antigen polyvinylpyrollidine (PVP), and determined the allotype and phenotype of responding cells (48). Responding cells were predominantly bone marrow derived but with the surface phenotype (CD5+, CD23−, CD43+, IgDlo, IgMhi) of B-1 cells. Since bone marrow only inefficiently reconstitutes the B-1 compartment, the

1 Feb 2002

10:15

270

AR

AR152-10.tex

BERLAND

¥

AR152-10.SGM

LaTeX2e(2001/05/10)

P1: GJC

WORTIS

responding cells were most likely B-2 cells that had been driven to assume the B-1 phenotype by activation with PVP. This raises the issue of the extent to which the B-1 compartment is composed of cells that arose in the adult in response to a TI-2 antigen.

Annu. Rev. Immunol. 2002.20:253-300. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

B-1 Cells and Autoimmunity That B-1 cells often produce autoreactive antibodies, although nonpathogenic and of a low affinity, has led to interest in these cells as possible sources of the highaffinity, pathogenic autoantibodies seen in polysystem autoimmune disease. Indeed, there are examples, in both human and mouse, of an association between autoimmune disease and B-1 cells. In humans, elevated numbers of B-1 cells have been reported in patients with Sjorgen’s syndrome (111) and rheumatoid arthritis (112). In mice, increased numbers of B-1 cells have been observed in a number of naturally occurring and genetically manipulated strains that develop autoimmune manifestations (see below). How might B-1 cells contribute to autoimmune disease? In several respects Per-C B-1 cells have a phenotype similar to that of anergic B cells. Like anergic B cells they do not flux Ca2+ or proliferate in response to BCR ligation (10). Furthermore, anergic B cells have recently been shown to express low levels of CD5, which helps to maintain anergy (39). In addition, both anergic B cells (113) and B-1 cells (114) have elevated nuclear levels of the transcription factor NFATc. It is therefore possible that the induction of the B-1 phenotype serves to tolerize B cells expressing certain specificities while still keeping them available for certain responses. Autoimmune disease could occur as a result of failure to induce the B-1 phenotype. Alternatively, susceptibility to autoimmune disease might result from diminished negative regulation of B-1 cells. As a result B-1 cells producing low-affinity autoantibodies would receive T cell help, enter germinal centers, class switch, undergo somatic mutation, and as a result of affinity maturation, produce highaffinity IgG autoantibodies. B-1 cells are able to respond to T cell help and switch class (115). Somatic mutations are not seen in peritoneal B-1a cells (13), but they are found in some human CD5+ cells (116). Of course, it is also possible that once B-1 cells are recruited into germinal centers they lose the B-1 phenotype, which makes it difficult to establish the phenotype of the antecedent cell. In addition to the possibility of producing pathogenic autoantibody, B-1 cells may contribute to autoimmune disease by presenting self-antigen to autoreactive T cells or by virtue of their ability to secrete IL-10. This is discussed more fully below. We now turn to an examination of several mouse models of autoimmunity in which B-1 cells play a role. AUTOIMMUNE HEMOLYTIC ANEMIA (AHA) IN NZB MICE NZB mice spontaneously develop autoimmune hemolytic anemia as a result of the production of anti–red blood cell (RBC) antibodies (117). Even before overt disease, these mice exhibit immune system aberrations, including an increase in the number of B-1 cells (118).

1 Feb 2002

10:15

AR

AR152-10.tex

AR152-10.SGM

LaTeX2e(2001/05/10)

P1: GJC

Annu. Rev. Immunol. 2002.20:253-300. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

B-1 CELLS AND CD5

271

Evidence that B-1 cells may play a causative role in development of disease comes from a series of studies by Honjo and colleagues using mice transgenic for the heavy and light chains of 4C8, an NZB-derived, pathogenic anti-RBC immunoglobulin (119). Expression of 4C8 on a C57/BL6 (nonautoimmune) background resulting in deletion or anergy of transgene-expressing B cells in the spleen. Normal numbers of transgene-expressing B-1 cells were found in the PerC (119). About 50% of 4C8-transgenic mice develop autoimmune hemolytic anemia when housed under conventional conditions. Mice housed under germ-free or SPF conditions remained healthy, suggesting that exposure to pathogens (or perhaps certain commensal organisms) is necessary for disease (120). In additional studies, oral administration of LPS induced peritoneal as well as lamina propria B cells to secrete IgM and resulted in anemia in previously SPF-maintained, nonanemic 4C8 transgenic mice (121). Exogenous IL-10 or IL-5, but not IL-4, also activated peritoneal B cells leading to disease (122). Intra-peritoneal (ip) injection of mouse RBC’s induced apoptosis of the transgene-expressing peritoneal B-1 cells. Thus, the survival of these autoreactive cells is dependent on their sequestration from antigen (123). Strikingly, injection of RBCs into the peritoneum resulted in recovery from anemia, indicating that B-1 cells or their descendants might be the source of the secreted 4C8 autoantibody (123). This possibility was directly tested by enumerating 4C8-secreting cells in the bone marrow, spleen, and peritoneum of transgenic mice with severe, moderate, or no anemia. 4C8-secreting cells were found only in the peritonea, and their frequency correlated with the severity of anemia (123). 4C8-secreting cells were reported (data not shown) to have the surface phenotype of B-1 cells (123). Another indication that B-1 cells are involved in the development of disease is the fact that F1 males produced by crosses between xid mice and mice transgenic for 4C8 failed to produce 4C8 antibody or to develop disease (121). Xid, or X-linked immunodeficiency, is a spontaneous mutation of the X-linked Bruton’s tyrosine kinase (Btk) gene, which prevents B-1 development (33). Interpretation of this result is complicated by the fact that the xid mutation also affects, although less dramatically, B-2 development and function (34). These results suggest that activation of autoreactive B-1 cells by pathogens or bystander effects could lead to autoimmunity. Interestingly, in 4C8 transgenic mice B-1 development is dependent on T cells because when the 4C8 transgenes are crossed onto a Rag−/− background very few peritoneal B cells develop (124). Transfer of fetal thymus or administration of IL-5, IL-10, or LPS rescued B-1 development and resulted in anemia (124). T cells are not required for B-1 development in nontransgenic mice. These studies provide compelling evidence that autoreactive B-1 cells, sequestered from antigen in the peritoneum, can be activated by pathogens or cytokines to secrete autoantibodies and thereby cause disease in a nonautoimmune strain of mouse. However, in nontransgenic mice, B-1 cells, although the predominant source of nonpathogenic autoantibodies, do not produce pathogenic autoantibodies. In fact, in young 4C8 transgenic mice there are few peritoneal B cells, suggesting that transgene-expressing B-1 cells (or their progenitors) are

1 Feb 2002

10:15

Annu. Rev. Immunol. 2002.20:253-300. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

272

AR

AR152-10.tex

BERLAND

¥

AR152-10.SGM

LaTeX2e(2001/05/10)

P1: GJC

WORTIS

efficiently deleted in these mice (123). Presumably, a small number of cells that escape deletion end up sequestered from antigen in the peritoneum where they expand. The question, then, is what role, if any, do B-1 cells play in pathogenesis of autoimmunity in nontransgenic, NZB mice. Support for the idea that B-1 cells do play a role is provided by experiments in which peritoneal B-1 cells are selectively eliminated by repeated intraperitoneal injection of H2O (125). This treatment prevents development of AHA (125). It selectively targets B-1 cells because peritoneal B-2 cells eliminated by hypotonic shock are replenished from bone marrow precursors, whereas B-1 cells are not. In NZB mice, B-1 cells could, as in 4C8 transgenic mice, actually produce pathogenic autoantibody. Alternatively, they may play an accessory role, for example by presenting antigen to T cells (126) or secreting IL-10 (127). MURINE LUPUS ERYTHEMATOSUS (SLE) IN NZB/NZW F1 (NZB/W) MICE F1 hybrids of NZB and NZW mice spontaneously develop an autoimmune syndrome similar to human systemic lupus erythematosus (SLE) (117, 128). The murine disease (like that in human) is more common in females and is characterized by the production of somatically mutated IgG autoantibodies with high-affinity for nuclear antigens such as dsDNA (129) and nucleosomes (130). The presence of these antibodies results in deposition of immune complexes in the kidneys, and the development of glomerular nephritis, proteinurea, and death due to kidney failure (128). Like NZB mice, (NZB X NZW)F1 mice have an enlarged B-1 compartment (118). Breeding the xid gene into (NZB X NZW)F1 mice prevents B-1 development and development of disease (131). Similarly, selective depletion of peritoneal B-1 cells by ip injection of H2O reduces the severity of disease (125). These results suggest a role for B-1 cells in the development of SLE in NZB/W mice. In MRL/lpr, another murine model of SLE, there is no increase in B-1 cells and adoptive transfer experiments indicate that B-2 cells, not B-1 cells, are necessary for disease pathogenesis (132). (NZB X NZW)F1 mice exhibit both T and B cell abnormalities. Yet, transfer into SCID mice of cultured pre-B cells derived from (NZB X NZW)F1 fetal liver was sufficient to generate many manifestations of SLE including renal disease (133). Since T cells do not develop from these selected donor cells, it appears that B-lineage intrinsic defects play a primary role in the pathogenesis of lupuslike disease in (NZB X NZW)F1 mice. It is not known specifically what role, if any, B-1 cells play in this process. One possibility is that they (or their derivatives) produce pathogenic autoantibody. Before the development of disease, both (NZB X NZW)F1 mice and SLE patients have large numbers of B cells spontaneously producing nonpathogenic, low-affinity, IgM anti-DNA antibody (134). Disease progression is believed to entail an antigen-driven switch in these cells to the production of pathogenic, high-affinity, IgG anti-DNA antibody (135–137). There is one report that most of the spontaneous anti-DNA IgM was produced by B-1 cells in (NZB X NZW)F1 mice (138). However, another study found that in

1 Feb 2002

10:15

AR

AR152-10.tex

AR152-10.SGM

LaTeX2e(2001/05/10)

P1: GJC

Annu. Rev. Immunol. 2002.20:253-300. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

B-1 CELLS AND CD5

273

NZB.H-2bm12 mice, which are similar to (NZB X NZW)F1, there was no difference in the frequency of splenic B-1 versus B-2 cells secreting IgM anti-DNA IgM (139). Consistent with this, yet another group showed that sorted (NZB X NZW)F1 splenic B-2 and peritoneal B-1 cells were equally competent to produce IgM or IgG anti-DNA when transferred into SCID mice along with (NZB X NZW)F1 T cells (140). Interestingly, transferred B cells from both sources produced equivalent levels of IgG anti-DNA and were able to cause disease. Three of three mice reconstituted with splenic B-2 cells and two of three mice reconstituted with peritoneal B-1 cells developed proteinurea (140). Leaving aside questions of the purity of the transferred cells, which was not rigorously controlled for, this suggests that (NZB X NZW)F1 B-1 cells are capable, although not uniquely so, of producing pathogenic autoantibody. Therefore, one aspect of (NZB X NZW)F1 disease susceptibility may be a breakdown in the mechanisms preventing B-1 cell hypermutation and affinity maturation. In NZM2410 mice, an autoimmune strain derived from (NZB X NZW)F1 mice, B-1 cells express high levels of the costimulatory molecules B7-1, B7-2, and CD24 in comparison to B-2 cells or to B-1 cells from C57/BL6 mice (126). B-1 cells from these mice, compared to B-2 cells, exhibited enhanced antigen-presenting ability to T cells in vitro (126). This might enable them to better elicit the cognate T cell help presumably necessary for somatic mutation. Unfortunately, the antigenpresenting capabilities of NZM2410 B-1 cells were not compared to those of B-1 cells from a nonautoimmune strain. In addition to being a source of pathogenic autoantibodies, B-1 cells could contribute in other ways to autoimmunity. Their enhanced antigen-presenting ability might better activate T cells to provide help to B-2 cells. Enhanced T cell activation could also contribute to disease by a mechanism independent of antibody production. Chan et al. reported that in the MRL/lpr model of SLE, B cells expressing mIgM, but unable to secrete Ig, can activate T cells resulting in renal cellular infiltrates and nephritis (141). In addition, in (NZB X NZW)F1 (but not C57/BL6) mice, a significant fraction of splenic CD1hi B cells have a B-1a phenotype (142). CD1 is able to present sugar and/or lipid antigens to some subsets of T cells (143, 144). Splenic T cells transgenic for a CD1-reactive T cell receptor induced a lupuslike disease when transferred into irradiated nude mice along with nude BM (145) Although the CD1hi B-1a cells present in (NZB X NZW)F1 spleen appear not to secrete pathogenic autoantibody (N. Baumgarth, personal communication), they may contribute to autoimmunity by activating CD1-restricted splenic T cells in (NZB X NZW)F1 mice. Finally, B-1 cells could contribute to SLE by their secretion of IL-10 (127). This interleukin is important in SLE pathogenesis. Continuous injection of anti-IL-10 antibody delayed the development of autoimmune disease in (NZB X NZW)F1 mice while continuous administration of IL-10 accelerated it (146). AUTOIMMUNITY IN GENE TARGETED AND TRANSGENIC MICE A number of gene knockout and transgenic mice have increased B-1 cell numbers and manifestations of autoimmune disease. Targeted disruption of the src family tyrosine kinase lyn

1 Feb 2002

10:15

Annu. Rev. Immunol. 2002.20:253-300. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

274

AR

AR152-10.tex

BERLAND

¥

AR152-10.SGM

LaTeX2e(2001/05/10)

P1: GJC

WORTIS

results in B cells that are hyperesponsive to BCR ligation because of impaired CD22 (147–149) and Fcγ RII (149–151) mediated downregulation of BCR signaling. Lyn knockout mice have large numbers of B-1 cells in their lymph nodes (151) and exhibit increasing levels of IgG autoantibodies and glomerular nephritis with age (149, 151, 152). The targeted disruption of the CD22 gene leads to enhanced BCR signaling (153–156) and, in one of four independently generated lines, to an increase in number of peritoneal B-1 cells (156). In another of the four, the number of peritoneal B-1 cells was not increased to a statistically significant extent, but the fraction of peritoneal cells with a B-1 phenotype was significantly increased (155). In the line found to have increased B-1 cell numbers, mice were followed for up to 20 months and found to develop high titers of IgG antibody specific for dsDNA, cardiolipin, and myeloperoxidase (157). These antibodies were of high affinity and somatically mutated (157). The mice did not develop other manifestations of autoimmune disease. Mice transgenic for the tumor necrosis factor (TNF)-ligand family member Blys (also termed TNF4, BAFF, TALL-1, and THANK), which is a costimulator of B cells, exhibit immune system dysregulation including increased serum IgM and IgG as well as the appearance in some older animals of anti-DNA antibodies, rheumatoid factor, and high levels of circulating immune complexes. Some animals exhibited Ig deposition in the kidneys and others some frank glomerular nephritis (158, 159). Transgenic lines were developed by two groups. One placed Blys under the control of a liver-specific promoter (158), while the other used a B cell specific promoter (159). In the latter line, but not the former, there was an increase in the number of splenic B-1 cells (159). The fact that autoimmune manifestations occur in Blys transgenic mice whether or not there was expansion of the B-1 compartment may indicate that B-1 cells are not responsible for the autoimmunity. It is possible, however, that there is B-1 dysregulation that does not consistently lead to an increase in B-1 number but is important in pathogenesis. Mice in which a TNFα transgene was introduced under the control of the lungspecific surfactant protein C promoter (SP-C/TNF) developed pulmonary fibrosis (160). Subsequently it was determined that lung interstitial mononuclear cells contain an elevated fraction of CD5+/IgM+ cells (161). This may be downstream of an increase in osteopontin (OPN). SP-C/TNF transgenic mice exhibit elevated levels of OPN mRNA in their lungs (160, 161); in a different study OPN transgenic mice were shown to have an increased fraction of PerC B-1a cells (162).

CD5+ ON B CELLS The transmembrane glycoprotein CD5 (historically, Lyt-1 or Ly-1 in mice) was first identified on T cells (163) by staining with a polyclonal antibody. Creation of a monoclonal antibody permitted the identification of CD5 on Thy-1− cells in

1 Feb 2002

10:15

AR

AR152-10.tex

AR152-10.SGM

LaTeX2e(2001/05/10)

P1: GJC

Annu. Rev. Immunol. 2002.20:253-300. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

B-1 CELLS AND CD5

275

the B cell areas of murine lymphoid tissue (164) and subsequently on a minor population of spleen cells with high levels of mIgM that was particularly evident in NZB mice (165–167). Fetal and neonatal tissues of normal mice were found to be rich in CD5+ B cells (168). At the same time Hayakawa and coworkers showed that CD5+ B cells had unique functions (72). Virtually parallel observations were made concerning a human lymphocyte surface antigen. A monoclonal antibody identified an antigen (Leu 1) on thymusdependent T cells and B cell–type chronic lymphatic leukemia cells (169). Others confirmed finding this 67-kDa protein on mature T cells (170) and normal adult B cells (171). As in mice, fetal and neonatal tissues were enriched for these cells (168, 172), and they had unique functions (173, 174). CD5 has now been identified and at least partially characterized in humans, mice, rabbits, rats, cattle, sheep, pigs, and chickens. The chicken CD5 (69 kDa with a 57-kDa protein core) is expressed on all T and B cells, although there are fewer molecules on the latter (175). Porcine CD5, of which only a small, highly (96%) homologous region has been sequenced, is expressed on T cells and a minor fraction of B cells (176). In cattle (177) and sheep (178) there appear to be minor populations of B cells that express surface CD5. All rabbit B cells are CD5+ (179). In the rat, CD5 is detected on T cells but is not observed on any B cells (180). In mice and humans the genes encoding CD5 and the related CD6 (181) are about 45 kb apart in a conserved syntenic group (182) that maps to mouse chromosome 19 (183) and human chromosome 11 (184, 185).

Expression As thymocytes mature, in particular as the amount of surface CD3 increases, the amount of CD5 also grows (186, 187). An in vitro experiment suggested that this could be a response to extracellular ligands: Murine CD4+CD8+ lymphoma cells responded to anti-CD3, concanavalin A, or phorbol esters plus ionomycin with a marked increase in surface CD5 (188). Double positive thymocytes also responded to TCR engagement by increasing their CD5 levels (189). Compelling in vivo evidence from several systems also shows that CD5 expression in thymocytes is driven by TCR binding of ligand (190). In a system employing mice transgenic for a TCR with anti–HY specificity, expression of the antigenic ligand in the context of the appropriate MHC class I caused thymocyte deletion. Dutz and coworkers went on to show that when β2-microglobulin was not present, H-Y expression did not delete the majority of CD4+ CD8+ thymocytes but did allow the development of thymocytes with an unusually low level of CD5 (191). These conclusions were confirmed (192–194) and further extended in an important paper by Azzam et al. (195) that showed that low level expression of CD5 on double negative cells did not require receptor gene expression as it was seen in Rag −/− mice. However, even this level of CD5 is likely to be CD3 receptor ligation–dependent, as CD3 is expressed on DN thymocytes of Rag −/− mice (196). Treatment in vivo with various amounts of anti-CD3 resulted in proportional

1 Feb 2002

10:15

Annu. Rev. Immunol. 2002.20:253-300. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

276

AR

AR152-10.tex

BERLAND

¥

AR152-10.SGM

LaTeX2e(2001/05/10)

P1: GJC

WORTIS

increases in CD5 expression as well as progression to the double positive stage. Lack of lck reduced but did not prevent CD5 expression. In their system, too, MHC-deficient mice expressed low levels of CD5. In toto their results show that the level of expression of the TCR, the avidity of ligation, and the strength of signal transduction (number of ITAMs in the CD3/TCR complex) collectively determined the amount of CD5 expressed. This supports a model in which the specificity of the TCR and the availability of appropriate antigen determine the avidity of TCR/ligand binding and thus the level of expressed CD5. In contrast, ligation of the TCR on mature peripheral CD4 or CD8 T cells does not seem to induce higher levels of CD5 expression (195). As CD5 decreases the sensitivity of the TCR (see below) its expression would alter the selected T cell repertoire. CD5 expression would block negative selection of some T cells with receptors that bind self-antigen with high avidity. At the same time CD5 would cause some low avidity cells to die from neglect. Conversely, loss of CD5 expression would allow deletion of some T cells that ordinarily bind selfantigen with insufficient avidity to induce tolerance. Loss of CD5 would also allow some low avidity self-reacting cells to be selected. These postulates were validated in experiments with TCR transgenic mice that co-expressed self antigens (196a). Similar evidence exists regarding B cell expression of CD5. The transformed murine pre-B cell line 70Z/3 expresses a low level of CD5. Expression is increased by treatment of the cells with LPS and decreased by IL-4 (197). CD5 expression can be induced by treatment of human peripheral B cells with phorbol esters (198, 199), and this can be blocked by simultaneous addition of IL-4 (200). In our hands murine CD5− B cells were not converted to CD5+ by this treatment (M. Teutsch, H. Wortis, unpublished). Ligation of the BCR of murine splenic B-2 cells by anti-IgM induces CD5 expression. In contrast, no induction occurs consequent to activation with LPS, T helper cells or CD40 ligand. In fact, CD40 ligand or IL-4 diminishes the induction of CD5 by anti-IgM (21, 201). Induction is at least in part at the level of transcription as it is blocked by actinomycin D and correlates with the accumulation of CD5 mRNA (24). In addition, CD5 50 flanking sequences allow anti-IgM treatment to induce expression of a luciferase reporter gene in transient transfection experiments (202). If anti-IgM mimics the ligation of the BCR by classic TI-2 antigens, then TI-2 antigens themselves should induce CD5 expression. This was tested by Whitmore, Haughton and Arnold who, after immunizing mice with the TI-2 antigen polyvinyl pyrrolidinone (PVP), used flow cytometry to sort CD5+ cells and CD5-cells (48). Similar regulation appears to occur in humans as incubation of CD5-tonsillar B cells with Staphlococcus aureus Cowan strain 1–induced CD5 expression (203). As in mice, IL-4 decreased CD5 expression on cells maintained in culture (204). PROMOTER Weichert et al. cloned a 1700-bp fragment 50 of the murine CD5 gene (205). By CAT assays they found that sequences between −125 and −27 were sufficient and necessary for transcriptional activity in transformed B and T cell

1 Feb 2002

10:15

AR

AR152-10.tex

AR152-10.SGM

LaTeX2e(2001/05/10)

P1: GJC

B-1 CELLS AND CD5

277

Annu. Rev. Immunol. 2002.20:253-300. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

lines. In contrast, the same CD5 flanking sequences could not drive expression in (nonlymphoid) NIH-3T3 cells. The region between −125 and −27 has no TATA box but does have potential Sp1, initiator and other recognizable elements. To date no elements in this proximal region have been identified by mutational analysis as important for B cell expression. Tung, Kunnavatana, Herzenberg & Herzenberg find that approximately 200 bp 50 of the translation start site there is a less than 50-bp region that is important for optimal expression of CD5 in transformed B and T lymphocytes. An ets binding site within this region appears to be essential for expression (205a).

Enhancer In order to identify regulatory sequences responsible for the induction of CD5 by BCR cross-linking, reporter constructs containing CD5 50 flanking sequences were introduced into murine splenic B cells. In all cases, reporter activity of unstimulated transfected cells was similar to background activity obtained by transfection with a promoterless reporter construct. A construct containing about 2200 bp 50 of the start site was induced about 10-fold by BCR ligation. Induction was abolished by deletion to −1965. Further analysis revealed that a 122-bp element comprised of sequences from −1919 to −2040 was sufficient to confer inducibility on the CD5 promoter proximal region (−6 to −277 relative to the ATG) even when placed in the reverse orientation and downstream of the reporter. Four complexes formed on this element after incubation with extracts from both induced and noninduced cells. Four additional complexes formed only in extracts from anti-IgM-treated B cells. Among these inducible complexes were two that were shown to contain NFAT. Point mutation of two putative NFAT binding sites severely compromised the activity of the enhancer in the transfection assay (202). This is consistent with the earlier observation that CD5 induction on splenic B cells was calcium dependent and cyclosporin A sensitive (201). A virtually identical element is now known to be located about 9.5-kb upstream of the CD5 gene in humans (cf. 114). Peritoneal CD5+ B cells have considerably more NFATc and in particular more nuclear NFATc than do their companion splenic B-2 cells. This raises the possibility that the constitutive expression of CD5 on B-1 cells may be NFAT-dependent. Preliminary data indicate that in NFATc knockout mice B-1a cell development and CD5 expression is compromised (R. Berland, H. Wortis, unpublished).

The CD5 Molecule EXTRACELLULAR DOMAINS CD5 is a type 1 transmembrane glycoprotein with a relative molecular mass of 67 kDa (206, 207). Comparison of the murine and human genes reveals 43% distal and 63% proximal identity in the extracellular region, and 90% identity for the intracellular carboxylterminal portion of the molecule (206). The extensive conservation of CD5, particularly of the cytoplasmic regions, suggests that it has important functions.

1 Feb 2002

10:15

Annu. Rev. Immunol. 2002.20:253-300. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

278

AR

AR152-10.tex

BERLAND

¥

AR152-10.SGM

LaTeX2e(2001/05/10)

P1: GJC

WORTIS

Partial characterization of the extracellular region of CD5 reveals it to be a rod-like structure with N-linked and sialylated O-linked glycans (208). It contains three scavenger receptor cysteine-rich (SRCR) domains (209, 210). CD5 and CD6 (see below) comprise a closely related group of cell surface molecules with three SRCR domains of about 110 amino acids (181). There is no known structural homology between the cytoplasmic regions of CD5, CD6, or any of the scavenger receptors. A third protein with a similar SRCR structure, Spα, is secreted by hematopoietic cells (181, 211). Spα was also independently cloned as AIM, an apoptosis inhibitor that is produced by macrophages (212, 213) and binds to unknown ligands on lymphocytes and macrophages. A natural soluble form of CD5 was recently identified in human serum (214). Soluble CD5, which may be generated by proteolytic cleavage, contains the extracellular SRCR domains of membrane-bound CD5. Factor I of the complement system is another secreted protein that contains an SRCR (215). Amino acid sequence comparison of SRCR domains allows them to be divided into two types, those with three (type A) or four (type B) disulfide bridges (216). Type B–containing proteins include CD5 and CD6 as well as Spα, CD163, and others variously expressed by lymphocytes, macrophages, and gut-associated tissues (181). Type A SRCR domains are found in Class A macrophage scavenger receptors and other proteins (cf 217). Physiological ligands of the type B SRCRs have not been established, with the notable exceptions of CD163, which binds haptoglobin-hemoglobin complexes (217a), and CD6, which binds (218) with ALCAM (activated leukocyte adhesion molecule) via its membrane-proximal SRCR domain (219, 220). Interestingly, CD5 and CD6 appear to differ at residues critical for binding to ALCAM (220a). Several cell surface molecules have been reported to be ligands of CD5. The first identified ligand was CD72 (221, 222). This was shown by a demonstration that CD5 purified from lysates of human or murine T cell lines bound to target B cells. This binding was specifically blocked by antibodies to CD72. However, Bikah and colleagues reported that a soluble CD5-Ig fusion protein bound to B cells and this binding was not blocked by either of two anti-CD72 antibodies. Nor did this fusion protein bind to L cells expressing transfected CD72 (223). Therefore, either CD5/CD72 is not a receptor/ligand pair or the fusion protein differed in ligand specificity from CD5 purified from membranes. Bikah reported that small resting B cells proliferated in response to the CD5 fusion protein, suggesting this receptor-ligand binding might have physiological significance. They named the putative ligand CD5L and reported that it was found only on B cells. It was just detectable on small resting cells but readily found on ex vivo peritoneal B cells and splenic B cells stimulated in vitro with LPS or anti-IgM. This may be the same ligand as described by Biancone et al., but there are differences in the characteristics of the partially purified proteins (10, 224). Evidence also suggests that a broadly expressed surface protein (225), as well as a region of the immunoglobulin VH framework (226), are CD5 ligands.

1 Feb 2002

10:15

AR

AR152-10.tex

AR152-10.SGM

LaTeX2e(2001/05/10)

P1: GJC

B-1 CELLS AND CD5

279

Annu. Rev. Immunol. 2002.20:253-300. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

To date, none of these has been shown to be a functional physiological ligand. As noted above, there is some evidence that CD5L can activate B cells. Also, a soluble recombinant form of CD5 acts in vitro as a modest costimulator of BCR-mediated proliferation, perhaps by binding to CD72 (227). Administration of a recombinant form of CD5 to mice with an antibody-induced membranous glomerulonephritis partially inhibited the disease (224), although the mechanism by which this effect is mediated has not been established. CYTOPLASMIC DOMAINS The five characterized mammalian (human, mouse, sheep, bovine, rat) CD5 molecules typically have cytoplasmic regions containing 96 amino acids. Of these, over 91 are identical or conserved. Each CD5 contains four tyrosines, four threonines, and eleven serines. This basic conservation is also preserved in the rabbit (C. Raman, K. L. Knight, unpublished). ASSOCIATION OF CD5 WITH TCR AND BCR Coimmunoprecipitation of Brij 96 lysates of human T cells revealed association of CD5 and CD3 with lck and fyn. Cross-linking of CD5 by antibody induced phosphorylation on the tyrosines of numerous proteins including CD5 itself. Treatment with anti-CD3 also caused phosphorylation of CD5 (228). Osman and colleagues (229) calculated that 10%– 20% of CD5 was CD3 associated. Fluorescence resonance energy transfer (FRET) demonstrated CD3 and CD5 to be within 10 nm of one another on human peripheral T cells (230). Based on co-immunoprecipitation of Brij lysates of normal human thymocytes, another group concluded that CD5, Zap-70, and CD3 were associated (231). CD phosphorylation correlated with the induced association with CD5. Human CD5 and IgM were co-immunoprecipitated from digitonin lysates of CD5+ T cells (232). This group also showed that CD5 is phosphorylated on tyrosine in response to CD5 ligation. In keeping with the idea that the BCR and CD5 are physically associated, CD5 and mIgM co-capped in response to cross-linking (233). CONSEQUENCES OF LIGATION: CANDIDATE EFFECTORS Within two minutes of ligation of human peripheral T cell CD3, CD5 is phosphorylated on tyrosines, threonines, and serines (234). In fact, as Pani and co-workers reported (235), the CD5 molecules in thymocytes and mature T cells have a basal level of phosphorylation on tyrosine that is increased by TCR cross-linking. They also found that a known negative regulator of activation, SH2-containing hematopoietic cell phosphatase (SHP-1), was associated with CD5 in both resting and TCR-activated cells. Perez-Villar (236) found that in the transformed human T cell line Jurkat, CD5 was basally phosphorylated on tyrosine and SHP-1 and CD5 could be coimmunoprecipitated (both ways) as could in vitro phosphatase activity. The association of CD5 and SHP-1 was not mediated by CD3 as a CD3− variant also showed this association. They failed to find CD5-associated SHP-2. TCR crosslinking increased both CD5 phosphorylation on tyrosine and the amount of

1 Feb 2002

10:15

Annu. Rev. Immunol. 2002.20:253-300. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

280

AR

AR152-10.tex

BERLAND

¥

AR152-10.SGM

LaTeX2e(2001/05/10)

P1: GJC

WORTIS

associated SHP-1. To explore the functional significance of these associations, stable transfectants expressing a fusion protein in which extracellular CD6 was combined with the transmembrane and cytoplasmic domains of CD5 were created. Cross-linking of the CD6/CD5 chimera with the TCR partially reduced calcium transients and phosphorylation on tyrosine of CD3, ZAP-70, Syk, and PLCγ 1 but did not affect tyrosine phosphorylation of lck. (However, it should be noted that this Jurkat line expressed endogenous, functional CD5.) Single Y > F mutations in the CD5 intracellular domain revealed that deletion of the juxta-membrane tyrosine was sufficient to ablate SHP-1 association, in vitro phosphatase activity, and the observed alterations in calcium transients and protein phosphorylation. Deletion mutations of the CD5 cytoplasmic region showed that a truncated protein with only the single membrane-proximal tyrosine retained all the observed negative regulatory activities and the ability to associate with SHP-1. The Perez-Villar study raises an interesting question about the mechanism by which CD5 might associate with the phosphatase SHP-1 and effect its activation. Current evidence indicates that optimal activation of SHP-1 requires engagement of both of its SH2s with P-TYRs (237). Yet, here mutations of CD5 show that phosphorylation of the juxta-membrane tyrosine is both necessary and sufficient for CD5 to mediate activation of SHP-1. It is known that loss of SHP-1, as in motheaten mice, results in T cells with increased sensitivity to activation by TCR cross-linking (235). Therefore, it is possible that SHP-1 is necessary for the negative regulatory activity of CD5. If so, either submaximal activation of SHP-1 via CD5 association is sufficient for negative regulation of T cells or association with CD5 may maximally activate SHP-1 despite the presence of only a single ITIM. Alternatively, the involvement of a third protein with a phosphorylated tyrosine may be involved. Conceivably, SHP-1 could bridge between CD5 and a phosphorylated tyrosine on a second protein or an adapter SH2 could bind with CD5 via a single SH2 and present two phosphorylated tyrosines to SHP-1. In B cells, CD5 negatively regulates signals through the BCR, as indicated by studies with CD5−/− mice. B-1 cells from these mice, in contrast to those from wild type, can flux Ca2+, activate NF-κB, and proliferate in response to BCR cross-linking. In the same study, B-1 cells from CD5+ mice were activated after BCR cross-linking if cells were first treated with biotinylated anti-CD5 and streptavidin. This was interpreted to be a consequence of sequestration of CD5 from other cell surface proteins, a likely candidate being the BCR complex itself (10). To examine the basis of the negative regulation of BCR signaling by CD5, a novel fusion protein containing the extracellular and transmembrane domains of Fcγ RIIB and the cytoplasmic region of CD5 without the juxta-membrane tyrosine was introduced into an FcR− variant of the transformed CD5-B cell line A20 (238). Co-cross-linking the BCR and this recombinant protein significantly decreased calcium transients, ERK2 activation, and the late but not the early phosphorylation of PLCγ 1. In contrast with other studies, anti-CD5 failed to co-immunoprecipitate SHIP, SHP-1, or SHP-2. Removal of the 16 aa segment Y429 through L444 ablated the negative regulation of the BCR by the CD5 recombinant protein. Taken

1 Feb 2002

10:15

AR

AR152-10.tex

AR152-10.SGM

LaTeX2e(2001/05/10)

P1: GJC

Annu. Rev. Immunol. 2002.20:253-300. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

B-1 CELLS AND CD5

281

together, these results suggest that CD5 may have an inhibitory element in addition to the juxta-membrane ITIM. Other experiments suggest that in B1a cells negative regulation by SHP-1 can be relieved by ligation and sequestration of CD5 (239). Using the two-hybrid system, two potential ligands of the CD5 cytoplasmic region were identified: CAM kinase IIδ and Tctex-1 (240). Tctex-1 is a component of the dynein motor complex. The association of CAM kinase IIδ and CD5 was further explored in vitro using fusion proteins. A 33 amino acid fragment containing residues of the juxta-membrane portion of the cytoplasmic region of CD5 bound to CAM kinase IIδ. Interestingly, this segment partly overlaps with a sequence that has homology with a CAM kinase IIδ association domain. CAM kinase II was previously shown to be activated through the BCR (241). Co-immunoprecipitation experiments failed to reveal any association of CD5 and CAM kinase IIδ in B cell lines in which both proteins are expressed. Nevertheless, it is attractive to think that CD5 might regulate the ability of CAM kinase II to phosphorylate calcineurin, thereby inhibiting the calciumdependent activation of proteins such as NFAT. Other yeast two-hybrid experiments suggested that casein kinase II (CK2) interacts with the cytoplasmic region of CD5. Co-immunoprecipitation of proteins from human B and T cell lines as well as mouse splenocytes revealed constitutive association of these two proteins. CD5 was phosphorylated in vitro by CK2. In B or T cell lines, cross-linking of CD5, but neither the B nor TCR, leads to the activation of CK2. A functional role for CK2 was not demonstrated (242, 243), and the significance of this protein-protein interaction remains to be established. It is intriguing that the cytoplasmic portion of CD163, a macrophage transmembrane protein with type B SRCRs, also associates with CK2 (243a).

FUNCTIONS OF CD5 Role of CD5 in B Cells In response to anti-TCR+ PMA or to concanavalin A, single positive thymocytes but not peripheral T cells from CD5 knockout mice proliferated more than wildtype cells. Yet, they proliferated equally well in response to PMA + ionomycin. Therefore, CD5 acts downstream of the TCR and, presumably, upstream of PLC. Thymocytes from CD5 knockout mice also show heightened Ca2+ transients and phosphorylation on the tyrosines of PLCγ 1 in response to TCR ligation. While candidate immediate targets were not identified, it is interesting to note that there appeared to be an alteration in the ratio of the phosphorylated isoforms of the ζ chain of CD3 and that there was markedly elevated tyrosine phosphorylation of Vav in the knockouts (245). Following TCR ligation there was phosphorylation on tyrosines of both the p21 and p23 isoforms of CD3. Previously it was observed that agonists induce similar levels of phosphorylation of CD3 p21 and p23, whereas anergizing agonists induce more phosphorylation of p21 than p23 (cf. 244). Interestingly, following treatment

1 Feb 2002

10:15

Annu. Rev. Immunol. 2002.20:253-300. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

282

AR

AR152-10.tex

BERLAND

¥

AR152-10.SGM

LaTeX2e(2001/05/10)

P1: GJC

WORTIS

with anti-TCR in the absence of CD5, phosphorylation of p23 was greater than that of p21. This suggests a model in which CD5 acts to alter a TCR ligation (agonist) signal toward anergy rather than activation. In the same study, absence of CD5 altered positive or negative selection of thymocytes in mice expressing any of three transgenic TCRs. That selection was variously affected suggests that CD5 acts to modify the strength of signal mediated through the TCR. Given that CD5 acts as a negative regulator, the loss of CD5 would alter the response to some ligands such that they would now be above the threshold of negative selection. Responses to other, weaker, antigen/TCR combinations would now be increased to the point that ligating cells would be positively selected (245, 246). In another system, in the presence of antibody to the TCR, antibody to CD5 provided a costimulatory stimulus that induced apoptosis of CD4+ CD8-medullary thymocytes (247). The ability of CD5 to act as a negative modifier of CD4 development by thymocyte selection was shown in a second transgenic TCR system. Here, the ability of CD5 to act as a negative regulator required the cytoplasmic portion of the molecule (248).

B Cell Function in CD5 Knockout Mice Studies of CD5 knockout mice showed that CD5 also acts as a negative regulator of B-1a cell activation (10). Purified B-2 and B-1 cells isolated from wild-type or CD5 knockout mice were stimulated with anti-IgM, anti-CD40, or LPS. As expected, B-1 cells from wild-type mice proliferated in response to CD40 or LPS but not anti-IgM. In contrast, B-1 cells from the knockout mice responded to all three. Calcium transients were sustained longer in the CD5− peritoneal B cells than in the wild type. Associated with these changes in signaling, CD5− B-1 cells showed reduced apoptosis and increased nuclear localization of NF-κB following BCR ligation. Therefore, in B-1a cells, as in T cells, CD5 is a negative regulator of antigen-induced signaling. The functional consequences of this negative regulatory activity of CD5 were examined by Hippen & Behrens (39). Close examination of anergic B cells generated in mice transgenic for antibody to hen’s egg lysozyme (HEL) as well as for the anergizing antigen soluble HEL (40) revealed that they expressed a very low level of CD5, even lower than seen on typical B-1a cells. This suggested that the induction of CD5 by autoantigen might be a mechanism by which the production of autoantibodies is avoided. This hypothesis was tested by breeding to produce CD5 knockout mice expressing the anti-HEL and soluble HEL transgenes. Consistent with this hypothesis, a fraction of the knockout, but none of the CD5+ wild-type mice, produced antibody to HEL.

Mice Transgenic for a B Cell Targeted CD5 Construct Chen, Matsura, & Kearney inserted CD5 cDNA into an expression construct such that it was under control of an immunoglobulin heavy chain promoter, the

1 Feb 2002

10:15

AR

AR152-10.tex

AR152-10.SGM

LaTeX2e(2001/05/10)

P1: GJC

Annu. Rev. Immunol. 2002.20:253-300. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

B-1 CELLS AND CD5

283

intronic enhancer, and portions of the switch region. With this construct they created transgenic mice that expressed high levels of CD5 on all B cells (249). In fact, CD5 was found on sIgM-negative, bone marrow cells, indicating that even pre-B and possibly pro-B cells might be capable of expressing CD5 on their surfaces. Additional preliminary experiments revealed that mature splenic CD5 transgenic mice did not differ from littermates in numbers of B cells or in the level of expression of B cell molecules mIgM, mIgD, MHC class II, CD38, or CD23. These B cells responded normally to a variety of mitogenic agents including antiIgM and CD40 ligand. There were no detectable differences in serum Ig levels in nonimmunized animals nor in responses to TI-2 or TD antigens. Notably, when immunized with NP-chicken gamma globulin, the transgenic mice produced anti-NP antibody. A limited number of these were sequenced and found to have somatic mutations at a frequency no less than seen in wild-type mice. This is evidence against the hypothesis that the infrequency of somatic mutation in the immunoglobulins produced by B-1a cells is a consequence of the expression of CD5 per se. The mutated sequences found in these NP-immunized transgenic mice were notable in that they were particularly enriched in a mutation, W33L, known to raise the binding affinity for NP tenfold (250). As CD5 is a negative regulator of BCRmediated activation, it may well be that only this higher affinity antibody can be selected in the presence of the transgenic CD5. That is, CD5 sets a higher threshold for BCR-mediated activation. Yet, in mice that were six months or older, there was a striking splenomegaly that could be attributed to a hyperplasia of marginal zone B cells and plasma cells. This was not a simple founder effect, as it was seen in animals derived from four founders. It could be that CD5 negatively regulates tolerogenic signaling in the MZ or that it acts positively by an unknown mechanism. ACKNOWLEDGMENT This work was supported by NIH grant AI R0115803. Visit the Annual Reviews home page at www.annualreviews.org

LITERATURE CITED 1. Vos Q, Lees A, Wu ZQ, Snapper CM, Mond JJ. 2000. B-cell activation by T-cellindependent type 2 antigens as an integral part of the humoral immune response to pathogenic microorganisms. Immunol. Rev. 176:154–70 2. Brodeur PH, Wortis HH. 1980. Regulation of thymus-independent responses: unresponsiveness to a second challenge of TNP-Ficoll is mediated by hapten-specific antibodies. J. Immunol. 125:1499–505

3. Zhang J, Liu YJ, MacLennan IC, Gray D, Lane PJ. 1988. B cell memory to thymusindependent antigens type 1 and type 2: the role of lipopolysaccharide in B memory induction. Eur. J. Immunol. 18:1417–24 4. Martin F, Kearney JF. 2000. B-cell subsets and the mature preimmune repertoire. Marginal zone and B1 B cells as part of a “natural immune memory.” Immunol. Rev. 175:70–79 5. Marcos MA, Huetz F, Pereira P, Andreu

1 Feb 2002

10:15

284

6.

Annu. Rev. Immunol. 2002.20:253-300. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

6a.

6b.

6c.

6d.

6e.

6f.

6g.

6h.

7.

8.

AR

AR152-10.tex

BERLAND

¥

AR152-10.SGM

LaTeX2e(2001/05/10)

P1: GJC

WORTIS

JL, Martinez AC, Coutinho A. 1989. Further evidence for coelomic-associated B lymphocytes. Eur. J. Immunol. 19:2031– 35 Kroese FG, Ammerlaan WA, Deenen GJ. 1992. Location and function of B-cell lineages. Ann. NY Acad. Sci. 651:44–58 Karras JG, Wang Z, Huo L, Howard RG, Frank DA, Rothstein TL. 1997. Signal transducer and activator of transcription-3 (STAT3) is constitutively activated in normal, self-renewing B-1 cells but only inducibly expressed in conventional B lymphocytes. J. Exp. Med. 185:1035–42 Rothstein TL, Fischer GM, Tanguay DA, Pavlovic S, Colarusso TP, Gerstein RM, Clarke SH, Chiles TC. 2000. STAT3 activation, chemokine receptor expression, and cyclin-Cdk function in B-1 cells. Curr. Top. Microbiol. Immunol. 252:121–30 Youinou P, Jamin C, Lydyard PM. 1999. CD5 expression in human B-cell populations. Immunol. Today 20:312–16 Bedoui S, Kuhlmann S, Nave H, Drube J, Pabst R, von Horsten S. 2001. Differential effects of neuropeptide Y (NPY) on leukocyte subsets in the blood: mobilization of B-1-like B-lymphocytes and activated monocytes. J. Neuroimmunol. 117:125–32 Knight KL, Winstead CR. 1997. B lymphocyte development in the rabbit. Int. Rev. Immunol. 15:129–63 Reynolds J. 1997. The genesis, tutelage and exodus of B cells in the ileal Peyer’s patch of sheep. Int. Rev. Immunol. 15:265–99 Meyer A, Parng CL, Hansal SA, Osborne BA, Goldsby RA. 1997. Immunoglobulin gene diversification in cattle. Int. Rev. Immunol. 15:165–83 Masteller EL, Pharr GT, Funk PE, Thompson CB. 1997. Avian B cell development. Int. Rev. Immunol. 15:185–267 Hayakawa K, Hardy RR. 1988. Normal, autoimmune, and malignant CD5+ B cells: the Ly-1 B lineage? Annu. Rev. Immunol. 6:197–218 Morris DL, Rothstein TL. 1993. Abnormal transcription factor induction through the

9.

10.

11.

12.

13.

14.

15.

16.

17.

18.

19.

surface immunoglobulin M receptor of B-1 lymphocytes. J. Exp. Med. 177:857– 61 Rothstein TL, Kolber DL. 1988. Anti-Ig antibody inhibits the phorbol ester-induced stimulation of peritoneal B cells. J. Immunol. 141:4089–93 Bikah G, Carey J, Ciallella JR, Tarakhovsky A, Bondada S. 1996. CD5-mediated negative regulation of antigen receptorinduced growth signals in B-1 B cells. Science 274:1906–9 Rothstein TL, Kolber DL. 1988. Peritoneal B cells respond to phorbol esters in the absence of co-mitogen. J. Immunol. 140:2880–85 Kantor AB, Merrill CE, Herzenberg LA, Hillson JL. 1997. An unbiased analysis of V(H)-D-J(H) sequences from B-1a, B1b, and conventional B cells. J. Immunol. 158:1175–86 Forster I, Gu H, Rajewsky K. 1988. Germline antibody V regions as determinants of clonal persistence and malignant growth in the B cell compartment. EMBO J. 7:3693– 703 Kantor AB, Herzenberg LA. 1993. Origin of murine B cell lineages. Annu. Rev. Immunol. 11:501–38 Hardy RR, Hayakawa K. 1994. CD5 B cells, a fetal B cell lineage. Adv. Immunol. 55:297–339 Herzenberg LA. 2000. B-1 cells: the lineage question revisited. Immunol. Rev. 175:9–22 Solvason N, Lehuen A, Kearney JF. 1991. An embryonic source of Ly1 but not conventional B cells. Int. Immunol. 3:543–50 Godin IE, Garcia-Porrero JA, Coutinho A, Dieterlen-Lievre F, Marcos MA. 1993. Para-aortic splanchnopleura from early mouse embryos contains B1a cell progenitors. Nature 364:67–70 Forster I, Rajewsky K. 1987. Expansion and functional activity of Ly-1+ B cells upon transfer of peritoneal cells into allotype-congenic, newborn mice. Eur. J. Immunol. 17:521–28

1 Feb 2002

10:15

AR

AR152-10.tex

AR152-10.SGM

LaTeX2e(2001/05/10)

P1: GJC

Annu. Rev. Immunol. 2002.20:253-300. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

B-1 CELLS AND CD5 20. Hayakawa K, Hardy RR, Stall AM, Herzenberg LA. 1986. Immunoglobulin-bearing B cells reconstitute and maintain the murine Ly-1 B cell lineage. Eur. J. Immunol. 16:1313–16 21. Cong YZ, Rabin E, Wortis HH. 1991. Treatment of murine CD5-B cells with anti-Ig, but not LPS, induces surface CD5: two B-cell activation pathways. Int. Immunol. 3:467–76 22. Haughton G, Arnold LW, Whitmore AC, Clarke SH. 1993. B-1 cells are made, not born. Immunol. Today 14:84–87; discussion 7–91 23. Clarke SH, Arnold LW. 1998. B-1 cell development: evidence for an uncommitted immunoglobulin (Ig)M+ B cell precursor in B-1 cell differentiation. J. Exp. Med. 187:1325–34 24. Wortis HH, Teutsch M, Higer M, Zheng J, Parker DC. 1995. B-cell activation by crosslinking of surface IgM or ligation of CD40 involves alternative signal pathways and results in different B-cell phenotypes. Proc. Natl. Acad. Sci. USA 92:3348–52 25. Rothstein TL, Kolber DL, Murphy TP, Cohen DP. 1991. Induction of phorbol ester responsiveness in conventional B cells after activation via surface Ig. J. Immunol. 147:3728–35 25a. Lutz C, Ledermann B, Kosco-Vilbois MH, Ochsenbein AF, Zinkernagel RM, Kohler G, Brombacher F. 1998. IgD can largely substitute for loss of IgM function in B cells. Nature 393:797–801 26. Chumley MJ, Dal Porto JM, Kawaguchi S, Cambier JC, Nemazee D, Hardy RR. 2000. A VH11V kappa 9 B cell antigen receptor drives generation of CD5+ B cells both in vivo and in vitro. J. Immunol. 164:4586–93 27. Hayakawa K, Asano M, Shinton SA, Gui M, Allman D, Stewart CL, Silver J, Hardy RR. 1999. Positive selection of natural autoreactive B cells. Science 285: 113–16 28. Hayakawa K, Carmack CE, Hyman R,

29.

30.

31.

32.

33.

34.

35.

36.

37.

285

Hardy RR. 1990. Natural autoantibodies to thymocytes: origin, VH genes, fine specificities, and the role of Thy-1 glycoprotein. J. Exp. Med. 172:869–78 Hardy RR, Li YS, Allman D, Asano M, Gui M, Hayakawa K. 2000. B-cell commitment, development and selection. Immunol. Rev. 175:23–32 Hayakawa K, Hardy RR. 2000. B-1 cell definition. Curr. Top. Microbiol. Immunol. 252:15–22 Pennell CA, Mercolino TJ, Grdina TA, Arnold LW, Haughton G, Clarke SH. 1989. Biased immunoglobulin variable region gene expression by Ly-1 B cells due to clonal selection. Eur. J. Immunol. 19:1289–95 Arnold LW, Pennell CA, McCray SK, Clarke SH. 1994. Development of B-1 cells: segregation of phosphatidyl choline-specific B cells to the B-1 population occurs after immunoglobulin gene expression. J. Exp. Med. 179:1585–95 Hayakawa K, Hardy RR, Herzenberg LA. 1986. Peritoneal Ly-1 B cells: genetic control, autoantibody production, increased lambda light chain expression. Eur. J. Immunol. 16:450–56 Satterthwaite AB, Witte ON. 2000. The role of Bruton’s tyrosine kinase in Bcell development and function: a genetic perspective. Immunol. Rev. 175:120– 27 Hayakawa K, Hardy RR. 2000. Development and function of B-1 cells. Curr. Opin. Immunol. 12:346–53 Arnold LW, McCray SK, Tatu C, Clarke SH. 2000. Identification of a precursor to phosphatidyl choline-specific B-1 cells suggesting that B-1 cells differentiate from splenic conventional B cells in vivo: cyclosporin A blocks differentiation to B-1. J. Immunol. 164:2924–30 Qian Y, Santiago C, Borrero M, Tedder T, Clarke SH. 2001. Lupus-specific antiribonucleoprotein B cell tolerance in nonautoimmune mice is maintained by differentiation to B-1 and governed by B cell

1 Feb 2002

10:15

286

38.

Annu. Rev. Immunol. 2002.20:253-300. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

39.

40.

41.

42.

43.

44.

45.

46.

47.

AR

AR152-10.tex

BERLAND

¥

AR152-10.SGM

LaTeX2e(2001/05/10)

P1: GJC

WORTIS

receptor signaling thresholds. J. Immunol. 166:2412–19 Lang J, Jackson M, Teyton L, Brunmark A, Kane K, Nemazee D. 1996. B cells are exquisitely sensitive to central tolerance and receptor editing induced by ultralow affinity, membrane-bound antigen. J. Exp. Med. 184:1685–97 Hippen KL, Tze LE, Behrens TW. 2000. CD5 maintains tolerance in anergic B cells. J. Exp. Med. 191:883–90 Goodnow CC, Crosbie J, Adelstein S, Lavoie TB, Smith-Gill SJ, Brink RA, Pritchard-Briscoe H, Wotherspoon JS, Loblay RH, Raphael K, et al. 1988. Altered immunoglobulin expression and functional silencing of self-reactive B lymphocytes in transgenic mice. Nature 334:676–82 Watanabe N, Nisitani S, Ikuta K, Suzuki M, Chiba T, Honjo T. 1999. Expression levels of B cell surface immunoglobulin regulate efficiency of allelic exclusion and size of autoreactive B-1 cell compartment. J. Exp. Med. 190:461–69 Lam KP, Rajewsky K. 1999. B cell antigen receptor specificity and surface density together determine B-1 versus B-2 cell development. J. Exp. Med. 190:471– 78 Herzenberg LA, Stall AM, Lalor PA, Sidman C, Moore WA, Parks DR. 1986. The Ly-1 B cell lineage. Immunol. Rev. 93:81– 102 Huang CA, Henry C, Iacomini J, Imanishi-Kari T, Wortis HH. 1996. Adult bone marrow contains precursors for CD5+ B cells. Eur. J. Immunol. 26:2537–40 Iacomini J, Imanishi-Kari T. 1992. The effect of an immunoglobulin mu transgene on B cell maturation. Eur. J. Immunol. 22:745–51 Thomas-Vaslin V, Coutinho A, Huetz F. 1992. Origin of CD5+ B cells and natural IgM-secreting cells: reconstitution potential of adult bone marrow, spleen and peritoneal cells. Eur. J. Immunol. 22:1243–51 Kantor AB, Stall AM, Adams S, Watan-

48.

49.

50.

51.

52.

53.

54.

55.

56.

57.

abe K, Herzenberg LA. 1995. De novo development and self-replenishment of B cells. Int. Immunol. 7:55–68 Whitmore AC, Haughton G, Arnold LW. 1996. Phenotype of B cells responding to the thymus-independent type-2 antigen polyvinyl pyrrolidinone. Int. Immunol. 8: 533–42 Stall AM, Wells SM, Lam KP. 1996. B-1 cells: unique origins and functions. Semin. Immunol. 8:45–59 Hardy RR, Hayakawa K. 1991. A developmental switch in B lymphopoiesis. Proc. Natl. Acad. Sci. USA 88:11550–54 Malynn BA, Yancopoulos GD, Barth JE, Bona CA, Alt FW. 1990. Biased expression of JH-proximal VH genes occurs in the newly generated repertoire of neonatal and adult mice. J. Exp. Med. 171:843–59 Sheehan KM, Brodeur PH. 1989. Molecular cloning of the primary IgH repertoire: a quantitative analysis of VH gene usage in adult mice. EMBO J. 8:2313–20 Gregoire KE, Goldschneider I, Barton RW, Bollum FJ. 1979. Ontogeny of terminal deoxynucleotidyl transferase-positive cells in lymphohemopoietic tissues of rat and mouse. J. Immunol. 123:1347–52 Li YS, Hayakawa K, Hardy RR. 1993. The regulated expression of B lineage associated genes during B cell differentiation in bone marrow and fetal liver. J. Exp. Med. 178:951–60 ten Boekel E, Melchers F, Rolink AG. 1997. Changes in the V(H) gene repertoire of developing precursor B lymphocytes in mouse bone marrow mediated by the preB cell receptor. Immunity 7:357–68 Melchers F, ten Boekel E, Seidl T, Kong XC, Yamagami T, Onishi K, Shimizu T, Rolink AG, Andersson J. 2000. Repertoire selection by pre-B-cell receptors and B-cell receptors, and genetic control of B-cell development from immature to mature B cells. Immunol. Rev. 175:33–46 Yancopoulos GD, Desiderio SV, Paskind M, Kearney JF, Baltimore D, Alt FW. 1984. Preferential utilization of the most

1 Feb 2002

10:15

AR

AR152-10.tex

AR152-10.SGM

LaTeX2e(2001/05/10)

P1: GJC

B-1 CELLS AND CD5

58.

Annu. Rev. Immunol. 2002.20:253-300. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

59.

60.

61.

62.

63.

64.

65.

65a.

JH-proximal VH gene segments in preB-cell lines. Nature 311:727–33 Marshall AJ, Wu GE, Paige GJ. 1996. Frequency of VH81x usage during B cell development: initial decline in usage is independent of Ig heavy chain cell surface expression. J. Immunol. 156:2077–84 Rolink A, Grawunder U, Winkler TH, Karasuyama H, Melchers F. 1994. IL-2 receptor alpha chain (CD25, TAC) expression defines a crucial stage in pre-B cell development. Int. Immunol. 6:1257–64 Martin F, Chen X, Kearney JF. 1997. Development of VH81X transgene-bearing B cells in fetus and adult: sites for expansion and deletion in conventional and CD5/B1 cells. Int. Immunol. 9:493–505 Marshall AJ, Doyen N, Bentolila LA, Paige CJ, Wu GE. 1998. Terminal deoxynucleotidyl transferase expression during neonatal life alters D(H) reading frame usage and Ig-receptor-dependent selection of V regions. J. Immunol. 161:6657–63 Wasserman R, Li YS, Shinton SA, Carmack CE, Manser T, Wiest DL, Hayakawa K, Hardy RR. 1998. A novel mechanism for B cell repertoire maturation based on response by B cell precursors to pre-B receptor assembly. J. Exp. Med. 187:259–64 Ceredig R, ten Boekel E, Rolink A, Melchers F, Andersson J. 1998. Fetal liver organ cultures allow the proliferative expansion of pre-B receptor-expressing preB-II cells and the differentiation of immature and mature B cells in vitro. Int. Immunol. 10:49–59 Ye J, McCray SK, Clarke SH. 1996. The transition of pre-BI to pre-BII cells is dependent on the VH structure of the mu/surrogate L chain receptor. EMBO J. 15:1524–33 Lalor PA, Herzenberg LA, Adams S, Stall AM. 1989. Feedback regulation of murine Ly-1 B cell development. Eur. J. Immunol. 19:507–13 Hitoshi Y, Yamaguchi N, Mita S, Sonoda E, Takaki S, Tominaga A, Takatsu K.

65b.

65c.

65d.

65e.

65f.

66.

67.

68.

287

1990. Distribution of IL-5 receptor-positive B cells. Expression of IL-5 receptor on Ly-1(CD5)+ B cells. J. Immunol. 144: 4218–25 Laabi Y, Strasser A. 2000. Immunology. Lymphocyte survival—ignorance is BLys. Science 289:883–84 Siegel RM, Lenardo MJ. 2001. To B or not to B: TNF family signaling in lymphocytes. Nat. Immunol. 2:577–78 Gross JA, Dillon SR, Mudri S, Johnston J, Littau A, Roque R, Rixon M, Schou O, Foley KP, Haugen H, McMillen S, Waggie K, Schreckhise RW, Shoemaker K, Vu T, Moore M, Grossman A, Clegg CH. 2001. TACI-Ig neutralizes molecules critical for B cell development and autoimmune disease impaired B cell maturation in mice lacking BLyS. Immunity 15:289– 302 Ishikawa S, Sato T, Abe M, Nagai S, Onai N, Yoneyama H, Zhang Y, Suzuki T, Hashimoto S, Shirai T, Lipp M, Matsushima K. 2001. Aberrant high expression of B lymphocyte chemokine (BLC/ CXCL13) by C11b+CD11c+ dendritic cells in murine lupus and preferential chemotaxis of B1 cells towards BLC. J. Exp. Med. 193:1393–402 Xiao G, Harhaj EW, Sun SC. 2001. NFkappaB-inducing kinase regulates the processing of NF-kappaB2 p100. Mol. Cell 7:401–9 Kroese FG, Butcher EC, Stall AM, Lalor PA, Adams S, Herzenberg LA. 1989. Many of the IgA producing plasma cells in murine gut are derived from selfreplenishing precursors in the peritoneal cavity. Int. Immunol. 1:75–84 Bos NA, Meeuwsen CG, Wostmann BS, Pleasants JR, Benner R. 1988. The influence of exogenous antigenic stimulation on the specificity repertoire of background immunoglobulin-secreting cells of different isotypes. Cell. Immunol. 112: 371–80 Haury M, Sundblad A, Grandien A, Barreau C, Coutinho A, Nobrega A. 1997.

1 Feb 2002

10:15

288

69.

Annu. Rev. Immunol. 2002.20:253-300. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

70.

71.

72.

73.

74.

75.

76.

77.

AR

AR152-10.tex

BERLAND

¥

AR152-10.SGM

LaTeX2e(2001/05/10)

P1: GJC

WORTIS

The repertoire of serum IgM in normal mice is largely independent of external antigenic contact. Eur. J. Immunol. 27:1557–63 Carroll MC, Prodeus AP. 1998. Linkages of innate and adaptive immunity. Curr. Opin. Immunol. 10:36–40 Gommerman JL, Carroll MC. 2000. Negative selection of B lymphocytes: a novel role for innate immunity. Immunol. Rev. 173:120–30 Masmoudi H, Mota-Santos T, Huetz F, Coutinho A, Cazenave PA. 1990. All T15 Id-positive antibodies (but not the majority of VHT15+ antibodies) are produced by peritoneal CD5+ B lymphocytes. Int. Immunol. 2:515–20 Hayakawa K, Hardy RR, Honda M, Herzenberg LA, Steinberg AD. 1984. Ly-1 B cells: functionally distinct lymphocytes that secrete IgM autoantibodies. Proc. Natl. Acad. Sci. USA 81:2494–98 Hardy RR, Carmack CE, Shinton SA, Riblet RJ, Hayakawa K. 1989. A single VH gene is utilized predominantly in antiBrMRBC hybridomas derived from purified Ly-1 B cells. Definition of the VH11 family. J. Immunol. 142:3643–51 Su SD, Ward MM, Apicella MA, Ward RE. 1991. The primary B cell response to the O/core region of bacterial lipopolysaccharide is restricted to the Ly-1 lineage. J. Immunol. 146:327–31 Baumgarth N, Herman OC, Jager GC, Brown L, Herzenberg LA. 1999. Innate and acquired humoral immunities to influenza virus are mediated by distinct arms of the immune system. Proc. Natl. Acad. Sci. USA 96:2250–55 Boes M, Esau C, Fischer MB, Schmidt T, Carroll M, Chen J. 1998. Enhanced B-1 cell development, but impaired IgG antibody responses in mice deficient in secreted IgM. J. Immunol. 160:4776–87 Ehrenstein MR, O’Keefe TL, Davies SL, Neuberger MS. 1998. Targeted gene disruption reveals a role for natural secretory IgM in the maturation of the primary

78.

79.

80.

81.

82.

83.

84.

85.

immune response. Proc. Natl. Acad. Sci. USA 95:10,089–93 Agenes F, Freitas AA. 1999. Transfer of small resting B cells into immunodeficient hosts results in the selection of a self-renewing activated B cell population. J. Exp. Med. 189:319–30 Briles DE, Nahm M, Schroer K, Davie J, Baker P, Kearney J, Barletta R. 1981. Antiphosphocholine antibodies found in normal mouse serum are protective against intravenous infection with type 3 streptococcus pneumoniae. J. Exp. Med. 153:694–705 Yother J, Forman C, Gray BM, Briles DE. 1982. Protection of mice from infection with Streptococcus pneumoniae by anti-phosphocholine antibody. Infect. Immun. 36:184–88 Shaw PX, Horkko S, Chang MK, Curtiss LK, Palinski W, Silverman GJ, Witztum JL. 2000. Natural antibodies with the T15 idiotype may act in atherosclerosis, apoptotic clearance, and protective immunity. J. Clin. Invest. 105:1731–40 Boes M, Prodeus AP, Schmidt T, Carroll MC, Chen J. 1998. A critical role of natural immunoglobulin M in immediate defense against systemic bacterial infection. J. Exp. Med. 188:2381–86 Reid RR, Prodeus AP, Khan W, Hsu T, Rosen FS, Carroll MC. 1997. Endotoxin shock in antibody-deficient mice: unraveling the role of natural antibody and complement in the clearance of lipopolysaccharide. J. Immunol. 159:970–75 Baumgarth N, Herman OC, Jager GC, Brown LE, Herzenberg LA, Chen J. 2000. B-1 and B-2 cell-derived immunoglobulin M antibodies are nonredundant components of the protective response to influenza virus infection. J. Exp. Med. 192:271–80 Fagarasan S, Shinkura R, Kamata T, Nogaki F, Ikuta K, Tashiro K, Honjo T. 2000. Alymphoplasia (aly)-type nuclear factor kappaB-inducing kinase (NIK) causes defects in secondary lymphoid

1 Feb 2002

10:15

AR

AR152-10.tex

AR152-10.SGM

LaTeX2e(2001/05/10)

P1: GJC

B-1 CELLS AND CD5

Annu. Rev. Immunol. 2002.20:253-300. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

86.

87.

88.

89.

90.

91.

92.

93.

tissue chemokine receptor signaling and homing of peritoneal cells to the gutassociated lymphatic tissue system. J. Exp. Med. 191:1477–86 Macpherson AJ, Gatto D, Sainsbury E, Harriman GR, Hengartner H, Zinkernagel RM. 2000. A primitive T cell-independent mechanism of intestinal mucosal IgA responses to commensal bacteria. Science 288:2222–26 Hooijkaas H, Benner R, Pleasants JR, Wostmann BS. 1984. Isotypes and specificities of immunoglobulins produced by germ-free mice fed chemically defined ultrafiltered “antigen-free” diet. Eur. J. Immunol. 14:1127–30 Hashimoto K, Handa H, Umehara K, Sasaki S. 1978. Germfree mice reared on an “antigen-free” diet. Lab. Anim. Sci. 28: 38–45 Pasparakis M, Alexopoulou L, Grell M, Pfizenmaier K, Bluethmann H, Kollias G. 1997. Peyer’s patch organogenesis is intact yet formation of B lymphocyte follicles is defective in peripheral lymphoid organs of mice deficient for tumor necrosis factor and its 55-kDa receptor. Proc. Natl. Acad. Sci. USA 94:6319–23 Neumann B, Luz A, Pfeffer K, Holzmann B. 1996. Defective Peyer’s patch organogenesis in mice lacking the 55-kD receptor for tumor necrosis factor. J. Exp. Med. 184:259–64 De Togni P, Goellner J, Ruddle NH, Streeter PR, Fick A, Mariathasan S, Smith SC, Carlson R, Shornick LP, Strauss-Schoenberger J, et al. 1994. Abnormal development of peripheral lymphoid organs in mice deficient in lymphotoxin. Science 264:703–7 Banks TA, Rouse BT, Kerley MK, Blair PJ, Godfrey VL, Kuklin NA, Bouley DM, Thomas J, Kanangat S, Mucenski ML. 1995. Lymphotoxin-alpha-deficient mice. Effects on secondary lymphoid organ development and humoral immune responsiveness. J. Immunol. 155:1685–93 Eugster HP, Muller M, Karrer U, Car BD,

94.

95.

96.

97.

98.

99.

100.

101. 102.

289

Schnyder B, Eng VM, Woerly G, Le Hir M, di Padova F, Aguet M, Zinkernagel R, Bluethmann H, Ryffel B. 1996. Multiple immune abnormalities in tumor necrosis factor and lymphotoxin-alpha doubledeficient mice. Int. Immunol. 8:23–36 Koni PA, Sacca R, Lawton P, Browning JL, Ruddle NH, Flavell RA. 1997. Distinct roles in lymphoid organogenesis for lymphotoxins alpha and beta revealed in lymphotoxin beta-deficient mice. Immunity 6:491–500 Rennert PD, Browning JL, Mebius R, Mackay F, Hochman PS. 1996. Surface lymphotoxin alpha/beta complex is required for the development of peripheral lymphoid organs. J. Exp. Med. 184:1999– 2006 Rennert PD, Browning JL, Hochman PS. 1997. Selective disruption of lymphotoxin ligands reveals a novel set of mucosal lymph nodes and unique effects on lymph node cellular organization. Int. Immunol. 9:1627–39 Yamamoto M, Rennert P, McGhee JR, Kweon MN, Yamamoto S, Dohi T, Otake S, Bluethmann H, Fujihashi K, Kiyono H. 2000. Alternate mucosal immune system: Organized Peyer’s patches are not required for IgA responses in the gastrointestinal tract. J. Immunol. 164:5184–91 Shinkura R, Kitada K, Matsuda F, Tashiro K, Ikuta K, Suzuki M, Kogishi K, Serikawa T, Honjo T. 1999. Alymphoplasia is caused by a point mutation in the mouse gene encoding Nf-kappa-binducing kinase. Nat. Genet. 22:74–77 Miyawaki S, Nakamura Y, Suzuka H, Koba M, Yasumizu R, Ikehara S, Shibata Y. 1994. A new mutation, aly, that induces a generalized lack of lymph nodes accompanied by immunodeficiency in mice. Eur. J. Immunol. 24:429–34 Bos NA, Cebra JJ, Kroese FGM. 2000. B1 cells and the intestinal microflora. Curr. Top. Microbiol. Immunol. 252:211–20 Deleted in proof Deleted in proof

1 Feb 2002

10:15

Annu. Rev. Immunol. 2002.20:253-300. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

290

AR

AR152-10.tex

BERLAND

¥

AR152-10.SGM

LaTeX2e(2001/05/10)

P1: GJC

WORTIS

103. Martin F, Oliver AM, Kearney JF. 2001. Marginal zone and B1 B cells unite in the early response against T- independent blood-borne particulate antigens. Immunity 14:617–29 104. Mond JJ, Lees A, Snapper CM. 1995. T cell-independent antigens type 2. Annu. Rev. Immunol. 13:655–92 105. Chang SP, Perlmutter RM, Brown M, Heusser CH, Hood L, Rittenberg MB. 1984. Immunologic memory to phosphocholine. IV. Hybridomas representative of Group I (T15-like) and Group II (nonT15-like) antibodies utilize distinct VH genes. J. Immunol. 132:1550–55 106. Cariappa A, Tang M, Parng C, Nebelitskiy E, Carroll M, Georgopoulos K, Pillai S. 2001. The follicular versus marginal zone B lymphocyte cell fate decision is regulated by Aiolos, Btk, and CD21. Immunity 14:603–15101 107. Guinamard R, Okigaki M, Schlessinger J, Ravetch JK. 2000. Absence of marginal zone B cells in Pyk-2-deficient mice defines their role in the humoral response. Nat. Immunol. 1:31–36 108. Oliver AM, Martin F, Gartland GL, Carter RH, Kearney JF. 1997. Marginal zone B cells exhibit unique activation, proliferative and immunoglobulin secretory responses. Eur. J. Immunol. 27:2366–74 109. Martin F, Kearney JF. 2000. Positive selection from newly formed to marginal zone B cells depends on the rate of clonal production, CD19, and btk. Immunity 12:39–49 110. Claflin JL, Lieberman R, Davie JM. 1974. Clonal nature of the immune response to phosphorylcholine. I. Specificity, class, and idiotype of phosphorylcholine-binding receptors on lymphoid cells. J. Exp. Med. 139:58–73 111. Dauphinee M, Tovar Z, Talal N. 1988. B cells expressing CD5 are increased in Sjogren’s syndrome. Arthritis Rheum. 31:642–47 112. Youinou P, Mackenzie L, Katsikis P, Merdrignac G, Isenberg DA, Tuaillon N,

113.

114.

115.

116.

117.

118.

119.

120.

121.

Lamour A, Le Goff P, Jouquan J, Drogou A, et al. 1990. The relationship between CD5-expressing B lymphocytes and serologic abnormalities in rheumatoid arthritis patients and their relatives. Arthritis Rheum. 33:339–48 Dolmetsch RE, Lewis RS, Goodnow CC, Healy JI. 1997. Differential activation of transcription factors induced by Ca2+ response amplitude and duration. Nature 386:855–58 Berland R, Wortis HH. 2000. Role of NFAT in the regulation of B-1 cells. Curr. Top. Microbiol. Immunol. 252:131–40 Taki S, Schmitt M, Tarlinton D, Forster I, Rajewsky K. 1992. T cell-dependent antibody production by Ly-1 B cells. Ann. NY Acad. Sci. 651:328–35 Ebeling SB, Schutte ME, Logtenberg T. 1993. The majority of human tonsillar CD5+ B cells express somatically mutated V kappa 4 genes. Eur. J. Immunol. 23:1405–8 Howie JB, Helyer BJ. 1968. The immunology and pathology of NZB mice. Adv. Immunol. 9:215–66 Hayakawa K, Hardy RR, Parks DR, Herzenberg LA. 1983. The “Ly-1 B” cell subpopulation in normal immunodefective, and autoimmune mice. J. Exp. Med. 157:202–18 Okamoto M, Murakami M, Shimizu A, Ozaki S, Tsubata T, Kumagai S, Honjo T. 1992. A transgenic model of autoimmune hemolytic anemia. J. Exp. Med. 175:71– 79 Murakami M, Nakajima K, Yamazaki K, Muraguchi T, Serikawa T, Honjo T. 1997. Effects of breeding environments on generation and activation of autoreactive B-1 cells in anti-red blood cell autoantibody transgenic mice. J. Exp. Med. 185:791– 94 Murakami M, Tsubata T, Shinkura R, Nisitani S, Okamoto M, Yoshioka H, Usui T, Miyawaki S, Honjo T. 1994. Oral administration of lipopolysaccharides activates B-1 cells in the peritoneal

1 Feb 2002

10:15

AR

AR152-10.tex

AR152-10.SGM

LaTeX2e(2001/05/10)

P1: GJC

B-1 CELLS AND CD5

Annu. Rev. Immunol. 2002.20:253-300. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

122.

123.

124.

125.

126.

127.

128.

129.

cavity and lamina propria of the gut and induces autoimmune symptoms in an autoantibody transgenic mouse. J. Exp. Med. 180:111–21 Nisitani S, Tsubata T, Murakami M, Honjo T. 1995. Administration of interleukin-5 or -10 activates peritoneal B-1 cells and induces autoimmune hemolytic anemia in anti-erythrocyte autoantibodytransgenic mice. Eur. J. Immunol. 25: 3047–52 Murakami M, Tsubata T, Okamoto M, Shimizu A, Kumagai S, Imura H, Honjo T. 1992. Antigen-induced apoptotic death of Ly-1 B cells responsible for autoimmune disease in transgenic mice. Nature 357:77–80 Nisitani S, Sakiyama T, Honjo T. 1998. Involvement of IL-10 in induction of autoimmune hemolytic anemia in antierythrocyte Ig transgenic mice. Int. Immunol. 10:1039–47 Murakami M, Yoshioka H, Shirai T, Tsubata T, Honjo T. 1995. Prevention of autoimmune symptoms in autoimmuneprone mice by elimination of B-1 cells. Int. Immunol. 7:877–82 Mohan C, Morel L, Yang P, Wakeland EK. 1998. Accumulation of splenic B1a cells with potent antigen-presenting capability in NZM2410 lupus-prone mice. Arthritis Rheum. 41:1652–62 O’Garra A, Chang R, Go N, Hastings R, Haughton G, Howard M. 1992. Ly-1 B (B-1) cells are the main source of B cellderived interleukin 10. Eur. J. Immunol. 22:711–17 Andrews BS, Eisenberg RA, Theofilopoulos AN, Izui S, Wilson CB, McConahey PJ, Murphy ED, Roths JB, Dixon FJ. 1978. Spontaneous murine lupus-like syndromes. Clinical and immunopathological manifestations in several strains. J. Exp. Med. 148:1198–215 Hirose S, Wakiya M, Kawano-Nishi Y, Yi J, Sanokawa R, Taki S, Shimamura T, Kishimoto T, Tsurui H, Nishimura H, et al. 1993. Somatic diversification and

130.

131.

132.

133.

134.

135.

136.

137.

138.

291

affinity maturation of IgM and IgG antiDNA antibodies in murine lupus. Eur. J. Immunol. 23:2813–20 Gioud M, Kotzin BL, Rubin RL, Joslin FG, Tan EM. 1983. In vivo and in vitro production of anti-histone antibodies in NZB/NZW mice. J. Immunol. 131:269– 74 Steinberg BJ, Smathers PA, Frederiksen K, Steinberg AD. 1982. Ability of the xid gene to prevent autoimmunity in (NZB X NZW)F1 mice during the course of their natural history, after polyclonal stimulation, or following immunization with DNA. J. Clin. Invest. 70:587–97 Reap EA, Sobel ES, Cohen PL, Eisenberg RA. 1993. Conventional B cells, not B-1 cells, are responsible for producing autoantibodies in lpr mice. J. Exp. Med. 177:69–78 Reininger L, Radaszkiewicz T, Kosco M, Melchers F, Rolink AG. 1992. Development of autoimmune disease in SCID mice populated with long-term “in vitro” proliferating (NZB × NZW)F1 pre-B cells. J. Exp. Med. 176:1343–53 Steward MW, Hay FC. 1976. Changes in immunoglobulin class and subclass of anti-DNA antibodies with increasing age in N/ZBW F1 hybrid mice. Clin. Exp. Immunol. 26:363–70 Marion TN, Bothwell AL, Briles DE, Janeway CA Jr. 1989. IgG anti-DNA autoantibodies within an individual autoimmune mouse are the products of clonal selection. J. Immunol. 142:4269–74 Marion TN, Tillman DM, Jou NT. 1990. Interclonal and intraclonal diversity among anti-DNA antibodies from an (NZB × NZW)F1 mouse. J. Immunol. 145:2322–32 Tillman DM, Jou NT, Hill RJ, Marion TN. 1992. Both IgM and IgG anti-DNA antibodies are the products of clonally selective B cell stimulation in (NZB × NZW)F1 mice. J. Exp. Med. 176:761– 79 Okada T, Abe M, Takiura F, Hirose S,

1 Feb 2002

10:15

292

Annu. Rev. Immunol. 2002.20:253-300. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

139.

140.

141.

142.

143.

144.

145.

146.

AR

AR152-10.tex

BERLAND

¥

AR152-10.SGM

LaTeX2e(2001/05/10)

P1: GJC

WORTIS

Shirai T. 1990. Distinct surface phenotypes of B cells responsible for spontaneous production of IgM and IgG antiDNA antibodies in autoimmune-prone NZB × NZW F1 mice. Autoimmunity 7:109–20 Klinman DM, Holmes KL, Conover J, Chiang BL, Gershwin ME. 1993. B-1a and conventional B cells from autoimmune NZB.H-2bm12 mice exhibit similar functional characteristics in vivo. Eur. J. Immunol. 23:1866–71 Ye YL, Chuang YH, Chiang BL. 1996. In vitro and in vivo functional analysis of CD5+ and CD5-B cells of autoimmune NZB x NZW F1 mice. Clin. Exp. Immunol. 106:253–58 Chan OT, Hannum LG, Haberman AM, Madaio MP, Shlomchik MJ. 1999. A novel mouse with B cells but lacking serum antibody reveals an antibody-independent role for B cells in murine lupus. J. Exp. Med. 189:1639–48 Zeng D, Lee MK, Tung J, Brendolan A, Strober S. 2000. Cutting edge: a role for CD1 in the pathogenesis of lupus in NZB/NZW mice. J. Immunol. 164:5000– 4 Beekman EM, Porcelli SA, Morita CT, Behar SM, Furlong ST, Brenner MB. 1994. Recognition of a lipid antigen by CD1-restricted alpha-beta+ T-cells. Nature 372:691–94 Sieling PA, Chatterjee D, Porcelli SA, Prigozy TI, Mazzaccaro RJ, Soriano T, Bloom BR, Brenner MB, Kronenberg M, Brennan PJ, et al. 1995. CD1-restricted T cell recognition of microbial lipoglycan antigens. Science 269:227–30 Zeng DF, Dick M, Cheng LR, Amano M, Dejbakhsh-Jones S, Huie P, Sibley R, Strober S. 1998. Subsets of transgenic T cells that recognize CD1 induce or prevent murine lupus: role of cytokines. J. Exp. Med. 187:525–36 Ishida H, Hastings R, Kearney J, Howard M. 1992. Continuous anti-interleukin 10 antibody administration depletes mice of

147.

148.

149.

150.

151.

152.

153.

154.

155.

Ly-1 B cells but not conventional B cells. J. Exp. Med. 175:1213–20 Cornall RJ, Cyster JG, Hibbs ML, Dunn AR, Otipoby KL, Clark EA, Goodnow CC. 1998. Polygenic autoimmune traits: Lyn, CD22, and SHP-1 are limiting elements of a biochemical pathway regulating BCR signaling and selection. Immunity 8:497–508 Smith KGC, Tarlinton DM, Doody GM, Hibbs ML, Fearon DT. 1998. Inhibition of the B cell by CD22: a requirement for Lyn. J. Exp. Med. 187:807–11 Nishizumi H, Horikawa K, MlinaricRascan I, Yamamoto T. 1998. A doubleedged kinase Lyn: a positive and negative regulator for antigen receptor-mediated signals. J. Exp. Med. 187:1343–48 Wang J, Koizumi T, Watanabe T. 1996. Altered antigen receptor signaling and impaired Fas-mediated apoptosis of B cells in Lyn-deficient mice. J. Exp. Med. 184:831–38 Chan VW, Meng F, Soriano P, DeFranco AL, Lowell CA. 1997. Characterization of the B lymphocyte populations in Lyndeficient mice and the role of Lyn in signal initiation and down-regulation. Immunity 7:69–81 Hibbs ML, Tarlinton DM, Armes J, Grail D, Hodgson G, Maglitto R, Stacker SA, Dunn AR. 1995. Multiple defects in the immune system of Lyn-deficient mice, culminating in autoimmune disease. Cell 83:301–11 Nitschke L, Carsetti R, Ocker B, Kohler G, Lamers MC. 1997. CD22 is a negative regulator of B-cell receptor signalling. Curr. Biol. 7:133–43 Otipoby KL, Andersson KB, Draves KE, Klaus SJ, Farr AG, Kerner JD, Perlmutter RM, Law CL, Clark EA. 1996. CD22 regulates thymus-independent responses and the lifespan of B cells. Nature 384:634–37 Sato S, Miller AS, Inaoki M, Bock CB, Jansen PJ, Tang ML, Tedder TF. 1996. CD22 is both a positive and negative regulator of B lymphocyte antigen receptor

1 Feb 2002

10:15

AR

AR152-10.tex

AR152-10.SGM

LaTeX2e(2001/05/10)

P1: GJC

B-1 CELLS AND CD5

156.

Annu. Rev. Immunol. 2002.20:253-300. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

157.

158.

159.

160.

161.

162.

signal transduction: altered signaling in CD22-deficient mice. Immunity 5:551–62 O’Keefe TL, Williams GT, Davies SL, Neuberger MS. 1996. Hyperresponsive B cells in CD22-deficient mice. Science 274:798–801 O’Keefe TL, Williams GT, Batista FD, Neuberger MS. 1999. Deficiency in CD22, a B cell-specific inhibitory receptor, is sufficient to predispose to development of high affinity autoantibodies. J. Exp. Med. 189:1307–13 Mackay F, Woodcock SA, Lawton P, Ambrose C, Baetscher M, Schneider P, Tschopp J, Browning JL. 1999. Mice transgenic for BAFF develop lymphocytic disorders along with autoimmune manifestations. J. Exp. Med. 190:1697–710 Gross JA, Johnston J, Mudri S, Enselman R, Dillon SR, Madden K, Xu W, ParrishNovak J, Foster D, Lofton-Day C, Moore M, Littau A, Grossman A, Haugen H, Foley K, Blumberg H, Harrison K, Kindsvogel W, Clegg CH. 2000. TACI and BCMA are receptors for a TNF homologue implicated in B-cell autoimmune disease. Nature 404:995–99 Miyazaki Y, Araki K, Vesin C, Garcia I, Kapanci Y, Whitsett JA, Piguet PF, Vassalli P. 1995. Expression of a tumor necrosis factor-alpha transgene in murine lung causes lymphocytic and fibrosing alveolitis. A mouse model of progressive pulmonary fibrosis. J. Clin. Invest. 96:250– 59 Nakama K, Miyazaki Y, Nasu M. 1998. Immunophenotyping of lymphocytes in the lung interstitium and expression of osteopontin and interleukin-2 mRNAs in two different murine models of pulmonary fibrosis. Exp. Lung Res. 24:57–70 Iizuka J, Katagiri Y, Tada N, Murakami M, Ikeda T, Sato M, Hirokawa K, Okada S, Hatano M, Tokuhisa T, Uede T. 1998. Introduction of an osteopontin gene confers the increase in B1 cell population and the production of anti-DNA autoantibodies. Lab. Invest. 78:1523–33

293

163. Boyse EA, Miyazawa M, Aoki T, Old LJ. 1968. Ly-A and Ly-B: two systems of lymphocyte isoantigens in the mouse. Proc. R. Soc. London Ser. B 170:175 164. Ledbetter JA, Herzenberg LA. 1979. Xenogeneic monoclonal antibodies to mouse lymphoid differentiation antigens. Immunol. Rev. 47:63–90 165. Okumura K, Hayakawa K, Tada T. 1982. Cell-to-cell interaction controlled by immunoglobulin genes. Role of Thy-1-, Lyt-1+, Ig+ (B0 ) cell in allotype-restricted antibody production. J. Exp. Med. 156:443–53 166. Manohar V, Brown E, Leiserson WM, Chused TM. 1982. Expression of Lyt-1 by a subset of B lymphocytes. J. Immunol. 129:532–38 167. Hardy RR, Hayakawa K, Parks DR, Herzenberg LA. 1983. Demonstration of B-cell maturation in X-linked immunodeficient mice by simultaneous three-colour immunofluorescence. Nature 306:270–72 168. Bofill M, Janossy G, Janossa M, Burford GD, Seymour GJ, Wernet P, Kelemen E. 1985. Human B cell development. II. Subpopulations in the human fetus. J. Immunol. 134:1531–38 169. Wang CY, Good RA, Ammirati P, Dymbort G, Evans RL. 1980. Identification of a p69,71 complex expressed on human T cells sharing determinants with B-type chronic lymphatic leukemic cells. J. Exp. Med. 151:1539–44 170. Engleman EG, Warnke R, Fox RI, Dilley J, Benike CJ, Levy R. 1981. Studies of a human T lymphocyte antigen recognized by a monoclonal antibody. Proc. Natl. Acad. Sci. USA 78:1791–95 171. Cappio FC, Novarino A, Matera L, Gavosto F. 1981. Cell surface markers and kinetic features in acute lymphoblastic leukaemia. Haematologica 66:419–25 172. Hardy RR, Hayakawa K. 1986. Development and physiology of Ly-1 B and its human homolog, Leu-1 B. Immunol. Rev. 93:53–79 173. Casali P, Burastero SE, Nakamura M,

1 Feb 2002

10:15

294

174.

Annu. Rev. Immunol. 2002.20:253-300. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

175.

176.

177.

178.

179.

180.

181.

182.

AR

AR152-10.tex

BERLAND

¥

AR152-10.SGM

LaTeX2e(2001/05/10)

P1: GJC

WORTIS

Inghirami G, Notkins AL. 1987. Human lymphocytes making rheumatoid factor and antibody to ssDNA belong to Leu-1+ B-cell subset. Science 236:77–81 Hardy RR, Hayakawa K, Shimizu M, Yamasaki K, Kishimoto T. 1987. Rheumatoid factor secretion from human Leu-1+ B cells. Science 236:81–83 Koskinen R, Gobel TW, Tregaskes CA, Young JR, Vainio O. 1998. The structure of avian CD5 implies a conserved function. J. Immunol. 160:4943–50 Appleyard GD, Wilkie BN. 1998. Porcine CD5 gene and gene product identified on the basis of inter-species conserved cytoplasmic domain sequences. Vet. Immunol. Immunopathol. 60:275–83 Yang TJ, Baldwin CL, Hanby-Flarida M, Mather JF, Rabinovsky ED. 1995. Monoclonal antibody BLT-1, specific for the bovine homologue of CD5, reacts with the majority of mature T cells, a subpopulation of B cells and stimulates T cell proliferation. Dev. Comp. Immunol. 19:337–46 Chevallier N, Berthelemy M, Laine V, Le Rhun D, Femenia F, Polack B, Naessens J, Levy D, Schwartz-Cornil I. 1998. B-1like cells exist in sheep. Characterization of their phenotype and behaviour. Immunology 95:178–84 Raman C, Knight KL. 1992. CD5+ B cells predominate in peripheral tissues of rabbit. J. Immunol. 149:3858–64 Vermeer LA, de Boer NK, Bucci C, Bos NA, Kroese FG, Alberti S. 1994. MRC OX19 recognizes the rat CD5 surface glycoprotein, but does not provide evidence for a population of CD5bright B cells. Eur. J. Immunol. 24:585–92 Aruffo A, Bowen MA, Patel DD, Haynes BF, Starling GC, Gebe JA, Bajorath J. 1997. CD6-ligand interactions: a paradigm for SRCR domain function? Immunol. Today 18:498–504 Lecomte O, Bock JB, Birren BW, Vollrath D, Parnes JR. 1996. Molecular linkage of the mouse CD5 and CD6 genes. Immunogenetics 44:385–90

183. Tada N, Kimura S, Hammerling U. 1982. Immunogenetics of mouse B-cell alloantigen systems defined by monoclonal antibodies and gene-cluster formation of these loci. Immunol. Rev. 69:99–126 184. Hecht BK, Kipps T, Johnston NK, Cannizaro LA. 1989. Leu-1 (CD5) cell surface antigen mapped to 11q13 by in situ hybridization. Cytogenet. Cell Genet. 51: 1012–19 185. Padilla O, Calvo J, Vila JM, Arman M, Gimferrer I, Places L, Arias MT, Pujana MA, Vives J, Lozano F. 2000. Genomic organization of the human CD5 gene. Immunogenetics 51:993–1001 186. Fowlkes BJ, Edison L, Mathieson BJ, Chused TM. 1985. Early T lymphocytes. Differentiation in vivo of adult intrathymic precursor cells. J. Exp. Med. 162:802– 22 187. Lanier LL, Allison JP, Phillips JH. 1986. Correlation of cell surface antigen expression on human thymocytes by multi-color flow cytometric analysis: implications for differentiation. J. Immunol. 137:2501–7 188. Richie ER, McEntire B, Phillips J, Allison JP. 1988. Altered expression of lymphocyte differentiation antigens on phorbol ester-activated CD4+8+ T cells. J. Immunol. 140:4115–22 189. Kearse KP, Takahama Y, Punt JA, Sharrow SO, Singer A. 1995. Early molecular events induced by T cell receptor (TCR) signaling in immature CD4+ CD8+ thymocytes: increased synthesis of TCRalpha protein is an early response to TCR signaling that compensates for TCR-alpha instability, improves TCR assembly, and parallels other indicators of positive selection. J. Exp. Med. 181:193–202 190. Punt JA, Osborne BA, Takahama Y, Sharrow SO, Singer A. 1994. Negative selection of CD4+CD8+ thymocytes by T cell receptor-induced apoptosis requires a costimulatory signal that can be provided by CD28. J. Exp. Med. 179:709–13 191. Dutz JP, Ong CJ, Marth J, Teh HS. 1995. Distinct differentiative stages of

1 Feb 2002

10:15

AR

AR152-10.tex

AR152-10.SGM

LaTeX2e(2001/05/10)

P1: GJC

B-1 CELLS AND CD5

Annu. Rev. Immunol. 2002.20:253-300. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

192.

193.

194.

195.

196.

196a.

197.

198.

CD4+CD8+ thymocyte development defined by the lack of coreceptor binding in positive selection. J. Immunol. 154:2588–99 Groves T, Katis P, Madden Z, Manickam K, Ramsden D, Wu G, Guidos CJ. 1995. In vitro maturation of clonal CD4+CD8+ cell lines in response to TCR engagement. J. Immunol. 154: 5011–22 Sheard M, Sugaya K, Furuta M, Tsuda S, Takahama Y. 1996. Heterogeneous expression of recombination activating genes and surface CD5 in CD3low CD4+ CD8+ thymocytes. Scand. J. Immunol. 43:619–25 Groves T, Parsons M, Miyamoto NG, Guidos CJ. 1997. TCR engagement of CD4+CD8+ thymocytes in vitro induces early aspects of positive selection, but not apoptosis. J. Immunol. 158:65–75 Azzam HS, Grinberg A, Lui K, Shen H, Shores EW, Love PE. 1998. CD5 expression is developmentally regulated by T cell receptor (TCR) signals and TCR avidity. J. Exp. Med. 188:2301–11 Wiest DL, Burgess WH, McKean D, Kearse KP, Singer A. 1995. The molecular chaperone calnexin is expressed on the surface of immature thymocytes in association with clonotype-independent CD3 complexes. EMBO J. 14:3425–33 Azzam HS, DeJarnette JB, Huang K, Emmons R, Park CS, Sommers CL, ElKhoury D, Shores EW, Love PE. 2001. Fine tuning of TCR signaling by CD5. J. Immunol. 166:5464–72 Jyonouchi H, Voss RM, Good RA. 1990. Up-regulation and down-regulation of cell surface and mRNA expression of CD5 antigen by various humoral factors on murine 70Z/3 pre-B cell leukemia cell line: IL-4 down-regulates CD5 antigen expression. Cell. Immunol. 130:66–78 Miller RA, Gralow J. 1984. The induction of Leu-1 antigen expression in human malignant and normal B cells by

199.

200.

201.

202.

203.

204.

205.

205a.

206.

295

phorbol myristic acetate (PMA). J. Immunol. 133:3408–14 Freedman AS, Freeman G, Whitman J, Segil J, Daley J, Levine H, Nadler LM. 1989. Expression and regulation of CD5 on in vitro activated human B cells. Eur. J. Immunol. 19:849–55 Zupo S, Dono M, Massara R, Taborelli G, Chiorazzi N, Ferrarini M. 1994. Expression of CD5 and CD38 by human CD5-B cells: requirement for special stimuli. Eur. J. Immunol. 24:1426– 33 Teutsch M, Higer M, Wang D, Wortis HH. 1995. Induction of CD5 on B and T cells is suppressed by cyclosporin A, FK-520 and rapamycin. Int. Immunol. 7:381–92 Berland R, Wortis HH. 1998. An NFATdependent enhancer is necessary for antiIgM-mediated induction of murine CD5 expression in primary splenic B cells. J. Immunol. 161:277–85 Morikawa K, Oseko F, Morikawa S. 1993. Induction of CD5 antigen on human CD5-B cells by stimulation with Staphylococcus aureus Cowan strain I. Int. Immunol. 5:809–16 Hidaka T, Kitani A, Hara M, Harigai M, Suzuki K, Kawaguchi Y, Ishizuka T, Kawagoe M, Nakamura H. 1992. IL-4 down-regulates the surface expression of CD5 on B cells and inhibits spontaneous immunoglobulin and IgMrheumatoid factor production in patients with rheumatoid arthritis. Clin. Exp. Immunol. 89:223–29 Weichert TR, Schwartz RC. 1995. Cloning of the murine CD5 promoter and its tissue-specific regulation. J. Immunol. 154:4603–12 Tung JW, Kunnavatana SS, Herzenberg LA. 2001. The regulation of CD5 expression in murine T cells. BMC Mol. Biol. 2:5 Huang HJ, Jones NH, Strominger JL, Herzenberg LA. 1987. Molecular cloning of Ly-1, a membrane glycoprotein of

1 Feb 2002

10:15

296

Annu. Rev. Immunol. 2002.20:253-300. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

207.

208.

209.

210.

211.

212.

213.

AR

AR152-10.tex

BERLAND

¥

AR152-10.SGM

LaTeX2e(2001/05/10)

P1: GJC

WORTIS

mouse T lymphocytes and a subset of B cells: molecular homology to its human counterpart Leu-1/T1 (CD5). Proc. Natl. Acad. Sci. USA 84:204–8 Jones NH, Clabby ML, Dialynas DP, Huang HJ, Herzenberg LA, Strominger JL. 1986. Isolation of complementary DNA clones encoding the human lymphocyte glycoprotein T1/Leu-1. Nature 323:346–49 McAlister MS, Brown MH, Willis AC, Rudd PM, Harvey DJ, Aplin R, Shotton DM, Dwek RA, Barclay AN, Driscoll PC. 1998. Structural analysis of the CD5 antigen-expression, disulphide bond analysis and physical characterisation of CD5 scavenger receptor super-family domain 1. Eur. J. Biochem. 257:131– 41 Freeman M, Ashkenas J, Rees DJ, Kingsley DM, Copeland NG, Jenkins NA, Krieger M. 1990. An ancient, highly conserved family of cysteinerich protein domains revealed by cloning type I and type II murine macrophage scavenger receptors. Proc. Natl. Acad. Sci. USA 87:8810–14 Aruffo A, Melnick MB, Linsley PS, Seed B. 1991. The lymphocyte glycoprotein CD6 contains a repeated domain structure characteristic of a new family of cell surface and secreted proteins. J. Exp. Med. 174:949–52 Gebe JA, Llewellyn M, Hoggatt H, Aruffo A. 2000. Molecular cloning, genomic organization and cell-binding characteristics of mouse Spalpha. Immunology 99:78–86 Miyazaki T, Hirokami Y, Matsuhashi N, Takatsuka H, Naito M. 1999. Increased susceptibility of thymocytes to apoptosis in mice lacking AIM, a novel murine macrophage-derived soluble factor belonging to the scavenger receptor cysteine-rich domain superfamily. J. Exp. Med. 189:413–22 Yusa S, Ohnishi S, Onodera T, Miyazaki T. 1999. AIM, a murine apoptosis

214.

215.

216.

217.

217a.

218.

219.

220.

inhibitory factor, induces strong and sustained growth inhibition of B lymphocytes in combination with TGF-beta1. Eur. J. Immunol. 29:1086–93 Gebe JA, Kiener PA, Ring HZ, Li X, Francke U, Aruffo A. 1997. Molecular cloning, mapping to human chromosome 1 q21-q23, and cell binding characteristics of Spalpha, a new member of the scavenger receptor cysteine-rich (SRCR) family of proteins. J. Biol. Chem. 272:6151–58 Perkins SJ, Smith KF, Sim RB. 1993. Molecular modelling of the domain structure of factor I of human complement by X-ray and neutron solution scattering. Biochem. J. 295:101–8 Resnick D, Pearson A, Krieger M. 1994. The SRCR superfamily: a family reminiscent of the Ig superfamily. Trends Biochem. Sci. 19:5–8 Hohenester E, Sasaki T, Timpl R. 1999. Crystal structure of a scavenger receptor cysteine-rich domain sheds light on an ancient superfamily. Nat. Struct. Biol. 6:228–32 Kristiansen M, Graversen JH, Jacobsen C, Sonne O, Hoffman HJ, Law SK, Moestrup SK. 2001. Identification of the haemoglobin scavenger receptor. Nature 409:198–201 Bowen MA, Bajorath J, Siadak AW, Modrell B, Malacko AR, Marquardt H, Nadler SG, Aruffo A. 1996. The aminoterminal immunoglobulin-like domain of activated leukocyte cell adhesion molecule binds specifically to the membrane-proximal scavenger receptor cysteine-rich domain of CD6 with a 1:1 stoichiometry. J. Biol. Chem. 271:17390–96 Bodian DL, Skonier JE, Bowen MA, Neubauer M, Siadak AW, Aruffo A, Bajorath J. 1997. Identification of residues in CD6 which are critical for ligand binding. Biochemistry 36:2637–41 Skonier JE, Bowen MA, Emswiler J, Aruffo A, Bajorath J. 1996. Recognition of diverse proteins by members of

1 Feb 2002

10:15

AR

AR152-10.tex

AR152-10.SGM

LaTeX2e(2001/05/10)

P1: GJC

B-1 CELLS AND CD5

220a.

Annu. Rev. Immunol. 2002.20:253-300. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

221.

222.

223.

224.

225.

226.

227.

228.

the immunoglobulin superfamily: delineation of the receptor binding site in the human CD6 ligand ALCAM. Biochemistry 35:12287–91 Hohenester E, Sasaki T, Timpl R. 1999. Crystal structure of a scavenger receptor cysteine-rich domain sheds light on an ancient superfamily. Nat. Struct. Biol. 6:228–32 Van de Velde H, von Hoegen I, Luo W, Parnes JR, Thielemans K. 1991. The Bcell surface protein CD72/Lyb-2 is the ligand for CD5. Nature 351:662–65 Luo W, Van de Velde H, von Hoegen I, Parnes JR, Thielemans K. 1992. Ly-1 (CD5), a membrane glycoprotein of mouse T lymphocytes and a subset of B cells, is a natural ligand of the B cell surface protein Lyb-2 (CD72). J. Immunol. 148:1630–34 Bikah G, Lynd FM, Aruffo AA, Ledbetter JA, Bondada S. 1998. A role for CD5 in cognate interactions between T cells and B cells, and identification of a novel ligand for CD5. Int. Immunol. 10:1185–96 Biancone L, Bowen MA, Lim A, Aruffo A, Andres G, Stamenkovic I. 1996. Identification of a novel inducible cellsurface ligand of CD5 on activated lymphocytes. J. Exp. Med. 184:811–19 Calvo J, Places L, Padilla O, Vila JM, Vives J, Bowen MA, Lozano F. 1999. Interaction of recombinant and natural soluble CD5 forms with an alternative cell surface ligand. Eur. J. Immunol. 29: 2119–29 Pospisil R, Fitts MG, Mage RG. 1996. CD5 is a potential selecting ligand for B cell surface immunoglobulin framework region sequences. J. Exp. Med. 184:1279–84 Van de Velde H, Thielemans K. 1996. Native soluble CD5 delivers a costimulatory signal to resting human B lymphocytes. Cell. Immunol. 172:84–91 Burgess KE, Yamamoto M, Prasad KV, Rudd CE. 1992. CD5 acts as a tyro-

229.

230.

231.

232.

233.

234.

235.

236.

297

sine kinase substrate within a receptor complex comprising T-cell receptor zeta chain/CD3 and protein-tyrosine kinases p56lck and p59fyn. Proc. Natl. Acad. Sci. USA 89:9311–15 Osman N, Ley SC, Crumpton MJ. 1992. Evidence for an association between the T cell receptor/CD3 antigen complex and the CD5 antigen in human T lymphocytes. Eur. J. Immunol. 22:2995–3000 Osman N, Lazarovits AI, Crumpton MJ. 1993. Physical association of CD5 and the T cell receptor/CD3 antigen complex on the surface of human T lymphocytes. Eur. J. Immunol. 23:1173–76 Gary-Gouy H, Lang V, Sarun S, Boumsell L, Bismuth G. 1997. In vivo association of CD5 with tyrosine-phosphorylated ZAP-70 and p21 phospho-zeta molecules in human CD3+ thymocytes. J. Immunol. 159:3739–47 Lankester AC, van Schijndel GM, Cordell JL, van Noesel CJ, van Lier RA. 1994. CD5 is associated with the human B cell antigen receptor complex. Eur. J. Immunol. 24:812–16 Jamin C, Lydyard PM, Le Corre R, Youinou PY. 1996. CD5+B cells: differential capping and modulation of IgM and CD5. Scand. J. Immunol. 43:73–80 Davies AA, Ley SC, Crumpton MJ. 1992. CD5 is phosphorylated on tyrosine after stimulation of the T-cell antigen receptor complex. Proc. Natl. Acad. Sci. USA 89:6368–72 Pani G, Fischer KD, Mlinaric-Rascan I, Siminovitch KA. 1996. Signaling capacity of the T cell antigen receptor is negatively regulated by the PTP1C tyrosine phosphatase. J. Exp. Med. 184:839–52 Perez-Villar JJ, Whitney GS, Bowen MA, Hewgill DH, Aruffo AA, Kanner SB. 1999. CD5 negatively regulates the T-cell antigen receptor signal transduction pathway: involvement of SH2containing phosphotyrosine phosphatase SHP-1. Mol. Cell. Biol. 19:2903– 12

1 Feb 2002

10:15

Annu. Rev. Immunol. 2002.20:253-300. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

298

AR

AR152-10.tex

BERLAND

¥

AR152-10.SGM

LaTeX2e(2001/05/10)

P1: GJC

WORTIS

237. Wang LL, Blasioli J, Plas DR, Thomas ML, Yokoyama WM. 1999. Specificity of the SH2 domains of SHP-1 in the interaction with the immunoreceptor tyrosine-based inhibitory motif-bearing receptor gp49B. J. Immunol. 162:1318– 23 238. Gary-Gouy H, Bruhns P, Schmitt C, Dalloul A, Daeron M, Bismuth G. 2000. The pseudo-immunoreceptor tyrosinebased activation motif of CD5 mediates its inhibitory action on B-cell receptor signaling. J. Biol. Chem. 275:548–56 239. Sen G, Bikah G, Venkataraman C, Bondada S. 1999. Negative regulation of antigen receptor-mediated signaling by constitutive association of CD5 with the SHP-1 protein tyrosine phosphatase in B-1 B cells. Eur. J. Immunol. 29:3319– 28 240. Bauch A, Campbell KS, Reth M. 1998. Interaction of the CD5 cytoplasmic domain with the Ca2+/calmodulin-dependent kinase IIdelta. Eur. J. Immunol. 28:2167–77 241. Valentine MA, Czernik AJ, Rachie N, Hidaka H, Fisher CL, Cambier JC, Bomsztyk K. 1995. Anti-immunoglobulin M activates nuclear calcium/calmodulin-dependent protein kinase II in human B lymphocytes. J. Exp. Med. 182: 1943–49 242. Raman C, Kuo A, Deshane J, Litchfield DW, Kimberly RP. 1998. Regulation of casein kinase 2 by direct interaction with cell surface receptor CD5. J. Biol. Chem. 273:19183–89 243. Calvo J, Vilda JM, Places L, Simarro M, Padilla O, Andreu D, Campbell KS, Aussel C, Lozano F. 1998. Human CD5 signaling and constitutive phosphorylation of C-terminal serine residues by casein kinase II. J. Immunol. 161:6022– 29 243a. Ritter M, Buechler C, Kapinsky M, Schmitz G. 2001. Interaction of CD163 with the regulatory subunit of casein kinase II (CKII) and dependence of CD163

244.

245.

246.

247.

248.

249.

250.

251.

252.

signaling on CKII and protein kinase C. Eur. J. Immunol. 31:999–1009 Madrenas J, Wange RL, Wang JL, Isakov N, Samelson LE, Germain RN. 1995. Zeta phosphorylation without ZAP-70 activation induced by TCR antagonists or partial agonists. Science 267:515–18 Tarakhovsky A, Muller W, Rajewsky K. 1994. Lymphocyte populations and immune responses in CD5-deficient mice. Eur. J. Immunol. 24:1678–84 Tarakhovsky A, Kanner SB, Hombach J, Ledbetter JA, Muller W, Killeen N, Rajewsky K. 1995. A role for CD5 in TCR-mediated signal transduction and thymocyte selection. Science 269:535– 37 Kishimoto H, Sprent J. 1999. Several different cell surface molecules control negative selection of medullary thymocytes. J. Exp. Med. 190:65–73 Pena-Rossi C, Zuckerman LA, Strong J, Kwan J, Ferris W, Chan S, Tarakhovsky A, Beyers AD, Killeen N. 1999. Negative regulation of CD4 lineage development and responses by CD5. J. Immunol. 163:6494–501 Chen X, Matsuura Y, Kearney JF. 1995. CD5 transgenic mice. Curr. Top. Microbiol. Immunol. 194:209–17 Weiss U, Rajewsky K. 1990. The repertoire of somatic antibody mutants accumulating in the memory compartment after primary immunization is restricted through affinity maturation and mirrors that expressed in the secondary response. J. Exp. Med. 172:1681–89 Khan WN, Alt FW, Gerstein RM, Malynn BA, Larsson I, Rathbun G, Davidson L, Muller S, Kantor AB, Herzenberg LA, et al. 1995. Defective B cell development and function in Btk-deficient mice. Immunity 3:283–99 Hendriks RW, de Bruijn MF, Maas A, Dingjan GM, Karis A, Grosveld F. 1996. Inactivation of Btk by insertion of lacZ reveals defects in B cell development

1 Feb 2002

10:15

AR

AR152-10.tex

AR152-10.SGM

LaTeX2e(2001/05/10)

P1: GJC

B-1 CELLS AND CD5

253.

Annu. Rev. Immunol. 2002.20:253-300. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

254.

255.

256.

257.

258.

259.

260.

only past the pre-B cell stage. EMBO J. 15:4862–72 Leitges M, Schmedt C, Guinamard R, Davoust J, Schaal S, Stabel S, Tarakhovsky A. 1996. Immunodeficiency in protein kinase c-beta-deficient mice. Science 273:788–91 Wang DM, Feng J, Wen RR, Marine JC, Sangster MY, Parganas E, Hoffmeyer A, Jackson CW, Cleveland JL, Murray PJ, Ihle JN. 2000. Phospholipase C-gamma2 is essential in the functions of Bcell and several Fc-receptors. Immunity 13:25–35 Hashimoto A, Takeda K, Inaba M, Sekimata M, Kaisho T, Ikehara S, Homma Y, Akira S, Kurosaki T. 2000. Cutting edge: essential role of phospholipase Cgamma 2 in B cell development and function. J. Immunol. 165:1738–42 Fruman DA, Snapper SB, Yballe CM, Davidson L, Yu JY, Alt FW, Cantley LC. 1999. Impaired B cell development and proliferation in absence of phosphoinositide 3-kinase p85-alpha. Science 283:393–97 Suzuki H, Terauchi Y, Fujiwara M, Aizawa S, Yazaki Y, Kadowaki T, Koyasu S. 1999. Xid-like immunodeficiency in mice with disruption of the p85-alpha subunit of phosphoinositide 3-kinase. Science 283:390–92 Engel P, Zhou LJ, Ord DC, Sato S, Koller B, Tedder TF. 1995. Abnormal B lymphocyte development, activation, and differentiation in mice that lack or overexpress the CD19 signal transduction molecule. Immunity 3:39–50 Rickert RC, Rajewsky K, Roes J. 1995. Impairment of T-cell-dependent B-cell responses and B-1 cell development in CD19-deficient mice. Nature 376:352– 55 Minegishi Y, Rohrer J, Coustan-Smith E, Lederman HM, Pappu R, Campana D, Chan AC, Conley ME. 1999. An essential role for BLNK in human B cell development. Science 286:1954–57

299

261. Pappu R, Cheng AM, Li B, Gong Q, Chiu C, Griffin N, White M, Sleckman BP, Chan AC. 1999. Requirement for B cell linker protein (BLNK) in B cell development. Science 286:1949–54 262. Jumaa H, Wollscheid B, Mitterer M, Wienands J, Reth M, Nielsen PJ. 1999. Abnormal development and function of B lymphocytes in mice deficient for the signaling adaptor protein SLP-65. Immunity 11:547–54 263. Hayashi K, Nittono R, Okamoto N, Tsuji S, Hara Y, Goitsuka R, Kitamura D. 2000. The B cell-restricted adaptor BASH is required for normal development and antigen receptor-mediated activation of B cells. Proc. Natl. Acad. Sci. USA 97:2755–60 264. Ahearn JM, Fischer MB, Croix D, Goerg S, Ma M, Xia J, Zhou X, Howard RG, Rothstein TL, Carroll MC. 1996. Disruption of the Cr2 locus results in a reduction in B-1a cells and in an impaired B cell response to T-dependent antigen. Immunity 4:251–62 265. Molina H, Holers VM, Li B, Fung Y, Mariathasan S, Goellner J, StraussSchoenberger J, Karr RW, Chaplin DD. 1996. Markedly impaired humoral immune response in mice deficient in complement receptors 1 and 2. Proc. Natl. Acad. Sci. USA 93:3357–61 266. Tarakhovsky A, Turner M, Schaal S, Mee PJ, Duddy LP, Rajewsky K, Tybulewicz VL. 1995. Defective antigen receptor-mediated proliferation of B and T cells in the absence of Vav. Nature 374:467–70 267. Zhang R, Alt FW, Davidson L, Orkin SH, Swat W. 1995. Defective signalling through the T- and B-cell antigen receptors in lymphoid cells lacking the vav proto-oncogene. Nature 374:470–73 268. Humbert PO, Corcoran LM. 1997. oct-2 gene disruption eliminates the peritoneal B-1 lymphocyte lineage and attenuates B-2 cell maturation and function. J. Immunol. 159:5273–84

1 Feb 2002

10:15

Annu. Rev. Immunol. 2002.20:253-300. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

300

AR

AR152-10.tex

BERLAND

¥

AR152-10.SGM

LaTeX2e(2001/05/10)

P1: GJC

WORTIS

269. Wang JH, Avitahl N, Cariappa A, Friedrich C, Ikeda T, Renold A, Andrikopoulos K, Liang L, Pillai S, Morgan BA, Georgopoulos K. 1998. Aiolos regulates B cell activation and maturation to effector state. Immunity 9:543–53 270. Kopf M, Brombacher F, Hodgkin PD, Ramsay AJ, Milbourne EA, Dai WJ, Ovington KS, Behm CA, Kohler G, Young IG, Matthaei KI. 1996. IL-5-deficient mice have a developmental defect in CD5+ B-1 cells and lack eosinophilia but have normal antibody and cytotoxic T cell responses. Immunity 4:15–24 271. Yoshida T, Ikuta K, Sugaya H, Maki K, Takagi M, Kanazawa H, Sunaga S, Kinashi T, Yoshimura K, Miyazaki J, Takaki S, Takatsu K. 1996. Defective B-1 cell development and impaired immunity against Angiostrongylus cantonensis in IL-5R alpha-deficient mice. Immunity 4:483–94 272. Hiroi T, Yanagita M, Iijima H, Iwatani K, Yoshida T, Takatsu K, Kiyono H. 1999. Deficiency of IL-5 receptor alphachain selectively influences the development of the common mucosal immune system independent IgA-producing B-1 cell in mucosa-associated tissues. J. Immunol. 162:821–28 273. Solvason N, Wu WW, Parry D, Mahony D, Lam EW, Glassford J, Klaus GG, Sicinski P, Weinberg R, Liu YJ, Howard M, Lees E. 2000. Cyclin D2 is essential for BCR-mediated proliferation and CD5 B cell development. Int. Immunol. 12:631–38 274. Sidman CL, Shultz LD, Hardy RR, Hayakawa K, Herzenberg LA. 1986. Production of immunoglobulin isotypes by Ly-1+ B cells in viable motheaten

275.

276.

277.

278.

279.

280.

281.

and normal mice. Science 232:1423– 25 Pan C, Baumgarth N, Parnes JR. 1999. CD72-deficient mice reveal nonredundant roles of CD72 in B cell development and activation. Immunity 11:495–506 Sato S, Ono N, Steeber DA, Pisetsky DS, Tedder TF. 1996. CD19 regulates B lymphocyte signaling thresholds critical for the development of B-1 lineage cells and autoimmunity. J. Immunol. 157:4371–78 Zhang L, Eddy A, Teng YT, Fritzler M, Kluppel M, Melet F, Bernstein A. 1995. An immunological renal disease in transgenic mice that overexpress Fli-1, a member of the ets family of transcription factor genes. Mol. Cell. Biol. 15:6961– 70 Tominaga A, Takaki S, Koyama N, Katoh S, Matsumoto R, Migita M, Hitoshi Y, Hosoya Y, Yamauchi S, Kanai Y, et al. 1991. Transgenic mice expressing a B cell growth and differentiation factor gene (interleukin 5) develop eosinophilia and autoantibody production. J. Exp. Med. 173:429–37 Vink A, Warnier G, Brombacher F, Renauld JC. 1999. Interleukin 9-induced in vivo expansion of the B-1 lymphocyte population. J. Exp. Med. 189:1413–23 Nemazee DA, Burki K. 1989. Clonal deletion of B lymphocytes in a transgenic mouse bearing anti-MHC class I antibody genes. Nature 337:562–66 Cyster JG, Goodnow CC. 1995. Protein tyrosine phosphatase 1C negatively regulates antigen receptor signaling in B lymphocytes and determines thresholds for negative selection. Immunity 2:13– 24

P1: FRK

February 1, 2002

12:47

Annual Reviews

AR152-FM

Annual Review of Immunology Volume 20, 2002

Annu. Rev. Immunol. 2002.20:253-300. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

CONTENTS FRONTISPIECE, Charles A. Janeway, Jr. A TRIP THROUGH MY LIFE WITH AN IMMUNOLOGICAL THEME, Charles A. Janeway, Jr.

xiv 1

THE B7 FAMILY OF LIGANDS AND ITS RECEPTORS: NEW PATHWAYS FOR COSTIMULATION AND INHIBITION OF IMMUNE RESPONSES, Beatriz M. Carreno and Mary Collins

29

MAP KINASES IN THE IMMUNE RESPONSE, Chen Dong, Roger J. Davis, and Richard A. Flavell

PROSPECTS FOR VACCINE PROTECTION AGAINST HIV-1 INFECTION AND AIDS, Norman L. Letvin, Dan H. Barouch, and David C. Montefiori T CELL RESPONSE IN EXPERIMENTAL AUTOIMMUNE ENCEPHALOMYELITIS (EAE): ROLE OF SELF AND CROSS-REACTIVE ANTIGENS IN SHAPING, TUNING, AND REGULATING THE AUTOPATHOGENIC T CELL REPERTOIRE, Vijay K. Kuchroo, Ana C. Anderson, Hanspeter Waldner, Markus Munder, Estelle Bettelli, and Lindsay B. Nicholson

55 73

101

NEUROENDOCRINE REGULATION OF IMMUNITY, Jeanette I. Webster, Leonardo Tonelli, and Esther M. Sternberg

125

MOLECULAR MECHANISM OF CLASS SWITCH RECOMBINATION: LINKAGE WITH SOMATIC HYPERMUTATION, Tasuku Honjo, Kazuo Kinoshita, and Masamichi Muramatsu

165

INNATE IMMUNE RECOGNITION, Charles A. Janeway, Jr. and Ruslan Medzhitov

KIR: DIVERSE, RAPIDLY EVOLVING RECEPTORS OF INNATE AND ADAPTIVE IMMUNITY, Carlos Vilches and Peter Parham ORIGINS AND FUNCTIONS OF B-1 CELLS WITH NOTES ON THE ROLE OF CD5, Robert Berland and Henry H. Wortis E PROTEIN FUNCTION IN LYMPHOCYTE DEVELOPMENT, Melanie W. Quong, William J. Romanow, and Cornelis Murre

197 217 253 301

LYMPHOCYTE-MEDIATED CYTOTOXICITY, John H. Russell and Timothy J. Ley

SIGNAL TRANSDUCTION MEDIATED BY THE T CELL ANTIGEN x

323

P1: FRK

Annu. Rev. Immunol. 2002.20:253-300. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

February 1, 2002

12:47

Annual Reviews

AR152-FM

CONTENTS

xi

RECEPTOR: THE ROLE OF ADAPTER PROTEINS, Lawrence E. Samelson INTERACTION OF HEAT SHOCK PROTEINS WITH PEPTIDES AND ANTIGEN PRESENTING CELLS: CHAPERONING OF THE INNATE AND ADAPTIVE IMMUNE RESPONSES, Pramod Srivastava CHROMATIN STRUCTURE AND GENE REGULATION IN THE IMMUNE SYSTEM, Stephen T. Smale and Amanda G. Fisher PRODUCING NATURE’S GENE-CHIPS: THE GENERATION OF PEPTIDES FOR DISPLAY BY MHC CLASS I MOLECULES, Nilabh Shastri, Susan

371

Schwab, and Thomas Serwold

395 427

463

THE IMMUNOLOGY OF MUCOSAL MODELS OF INFLAMMATION, Warren Strober, Ivan J. Fuss, and Richard S. Blumberg T CELL MEMORY, Jonathan Sprent and Charles D. Surh

GENETIC DISSECTION OF IMMUNITY TO MYCOBACTERIA: THE HUMAN MODEL, Jean-Laurent Casanova and Laurent Abel ANTIGEN PRESENTATION AND T CELL STIMULATION BY DENDRITIC CELLS, Pierre Guermonprez, Jenny Valladeau, Laurence Zitvogel, Clotilde Th´ery, and Sebastian Amigorena

495 551 581

621

NEGATIVE REGULATION OF IMMUNORECEPTOR SIGNALING, Andr´e Veillette, Sylvain Latour, and Dominique Davidson

669

CPG MOTIFS IN BACTERIAL DNA AND THEIR IMMUNE EFFECTS, Arthur M. Krieg

PROTEIN KINASE Cθ

709 IN

T CELL ACTIVATION, Noah Isakov and Amnon

Altman

RANK-L AND RANK: T CELLS, BONE LOSS, AND MAMMALIAN EVOLUTION, Lars E. Theill, William J. Boyle, and Josef M. Penninger PHAGOCYTOSIS OF MICROBES: COMPLEXITY IN ACTION, David M. Underhill and Adrian Ozinsky

761 795 825

STRUCTURE AND FUNCTION OF NATURAL KILLER CELL RECEPTORS: MULTIPLE MOLECULAR SOLUTIONS TO SELF, NONSELF DISCRIMINATION, Kannan Natarajan, Nazzareno Dimasi, Jian Wang, Roy A. Mariuzza, and David H. Margulies

853

INDEXES Subject Index Cumulative Index of Contributing Authors, Volumes 1–20 Cumulative Index of Chapter Titles, Volumes 1–20

ERRATA An online log of corrections to Annual Review of Immunology chapters (if any, 1997 to the present) may be found at http://immunol.annualreviews.org/

887 915 925

8 Feb 2002

9:52

AR

AR152-11.tex

AR152-11.SGM

LaTeX2e(2001/05/10) P1: GJC 10.1146/annurev.immunol.20.092501.162048

Annu. Rev. Immunol. 2002. 20:301–22 DOI: 10.1146/annurev.immunol.20.092501.162048 c 2002 by Annual Reviews. All rights reserved Copyright °

Annu. Rev. Immunol. 2002.20:301-322. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

E PROTEIN FUNCTION IN LYMPHOCYTE DEVELOPMENT Melanie W. Quong∗ , William J. Romanow∗ and Cornelis Murreˆ University of California, San Diego, Division of Biology, 9500 Gilman Drive, MC 0366, La Jolla, California 92093-0366; e-mail: [email protected], [email protected], [email protected]

Key Words E2A, Id, thymocyte, B lymphocyte, helix-loop-helix (HLH) ■ Abstract Lymphocytes arise from hematopoietic stem cells through the coordinated action of transcription factors. The E proteins (E12, E47, HEB and E2-2) have emerged as key regulators of both B and T lymphocyte differentiation. This review summarizes the current data and examines the various functions of E proteins and their antagonists, Id2 and Id3, throughout lymphoid maturation. Beyond an established role in B and T lineage commitment, E proteins continue to be essential at subsequent stages of development. E protein activity regulates the expression of surrogate and antigen receptor genes, promotes Ig and TCR rearrangements, and coordinates cell survival and proliferation with developmental progression in response to TCR signaling. Finally, this review also discusses the role of E47 as a tumor suppressor.

INTRODUCTION Self-renewing, hematopoietic stem cells (HSC) are multipotent, capable of generating erythroid, myeloid, dendritic, and lymphoid cell lineages. Lymphocyte development occurs through a common lymphoid progenitor (CLP), which has restricted lineage potential (1). CLPs develop into three distinct cell types: B and T lymphocytes and NK cells. Beginning with lineage commitment and continuing throughout maturation, both B and T lymphocytes develop through stages that can be defined by the rearrangement of antigen receptor genes, the acquisition or loss of cell surface and intracellular proteins, and responses to growth and survival factors. Gene expression that occurs during lymphocyte development is controlled, at least in part, by a class of transcription factors known as the helix-loop-helix ∗

These authors contributed equally to this manuscript. ˆ To whom correspondence should be addressed.

0732-0582/02/0407-0301$14.00

301

1 Feb 2002

10:17

Annu. Rev. Immunol. 2002.20:301-322. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

302

AR

QUONG

AR152-11.tex

¥

AR152-11.SGM

ROMANOW

¥

LaTeX2e(2001/05/10)

P1: GJC

MURRE

proteins (HLH). This review focuses on the roles of the class I HLH proteins and their antagonists in lymphoid development. Class I HLH proteins, also known as E proteins, were first identified based on their ability to bind with relative high affinity to the palindromic DNA sequence CANNTG, referred to as an E box site (2–4). E boxes have since been found in the promoter and enhancer regions of a wide variety of B and T lineagespecific genes. These include the enhancers in the immunoglobulin (Ig) loci and the T cell receptor (TCR) α and β loci, the promoters of the mb-1, λ5, and pre-Tα genes, and the CD4 silencer and enhancer elements (5–10). The E protein family includes five members, designated as E12, E47, HEB, and E2-2, in vertebrates, and the Drosophila gene product, daughterless (for review, see 11). They share a highly conserved motif termed the helix-loop-helix (HLH), which consists of two amphipathic α-helices separated by a loop structure (12). This domain imparts dimerization capabilities between HLH members (12). Immediately amino-terminal to the HLH is a conserved basic region that allows HLH dimers to bind to DNA. In addition to the HLH motif, the mammalian E proteins share two conserved transcriptional activation domains referred to as the AD1 and loop-helix (LH) domains (13–15). The AD1 domain forms an acidic helical region and is also present in Rtg3p, an HLH protein in S. cerevisiae (15, 16). The LH domain contains a potential loop located adjacent to a putative amphipathic helical structure and is conserved among all E protein members, including daughterless (13). The conservation of the activation domains suggests that the target molecules with which E proteins interact are conserved throughout evolution (11). Ablation of either activation region, through deletions or structural mutations, severely affects the transcriptional capabilities of the E proteins (13, 15). The E2A gene encodes for two E proteins, E12 and E47, which arise through differential splicing of the exon encoding for the HLH domain (4, 17). HEB and E2-2 are encoded by separate genes (3, 18). All of the E proteins have the ability to bind DNA as homo- or heterodimers (19). Although not ubiquitous, E proteins are broadly expressed; however, certain E protein complexes are restricted to specific cell types. In B lineage cells, the predominant E box binding complex is comprised of E47 homodimers, whereas in thymocytes, E47/HEB heterodimers predominate (9, 20–23). Additional control of E protein activity occurs through interaction with the Id HLH proteins (24). Members of this group lack the basic region and thus do not have the ability to bind to DNA (25). Id proteins efficiently heterodimerize with E protein members and effectively act as dominant-negative HLH proteins. Four mammalian members have been identified: Id1, Id2, Id3 and Id4, as well as one Drosophila homologue, extramacrochaete (emc) (25–31). Of the four Id proteins, only Id2 and Id3 are abundantly expressed in B and T lymphocytes (32, 33). Since the Id proteins can act as dominant-negative HLH proteins, the ratio of E proteins to Id proteins ultimately determines the level of E protein activity.

1 Feb 2002

10:17

AR

AR152-11.tex

AR152-11.SGM

LaTeX2e(2001/05/10)

P1: GJC

E PROTEINS IN LYMPHOID DEVELOPMENT

303

THE ROLES OF E PROTEINS IN B LYMPHOCYTE DEVELOPMENT

Annu. Rev. Immunol. 2002.20:301-322. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

B Lineage Commitment and Early Stages Commitment and progression through early stages of B lineage development is dependent on specific gene expression programs as well as environmental cues (Figure 1). The study of gene transcription has included regulatory elements of the Ig genes, as well as genes encoding the structural components of the B cell receptor (BCR), including Igα, Igβ, λ5, and VpreB. Additionally, components of the recombination machinery, namely RAG1 and RAG2, have been examined in great detail (34, 35). The analysis of genetically altered mice has led to the identification of a number of transcription factors, including E2A, EBF, and Pax5, that regulate many of these B lineage specific genes (36–39). E2A, EBF, and Pax5 act in a cascade and in synergy to establish and maintain the expression of target genes that promote the B lymphocyte maturation program (10). E2A-deficient mice display a complete block in B lineage development at a very early stage prior to the onset of IgH DJ rearrangement (Figure 2a) (36, 37).

Figure 1 Schematic diagram of murine B cell development. Successive developmental stages are shown together with their characteristic cell surface markers. Transcription factors indicated above the arrows denote the approximate developmental stages at which they have been described to function. The appropriate stages when Ig rearrangements are incurred are indicated by the horizontal lines. Vertical dashed lines delineate checkpoints requiring pre-BCR and BCR-mediated signals for further developmental progression.

1 Feb 2002

10:17

Annu. Rev. Immunol. 2002.20:301-322. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

304

AR

QUONG

AR152-11.tex

¥

AR152-11.SGM

ROMANOW

¥

LaTeX2e(2001/05/10)

P1: GJC

MURRE

Figure 2 Summary of the stages at which E proteins affect B and T lymphocyte development. (a) B cell development. E proteins have been shown to be required at several stages of development as indicated by horizontal arrows on the left side of the figure. The corresponding functions are described on the right side (see text for details). Cell surface expression of the antigen receptors are indicated where appropriate. At the commitment stage, E2-2 and HEB are required for generating wild-type levels of pro-B cells. (b) T cell development. Horizontal arrows on the left indicate where E proteins are required for progression. Blunt-ended arrows denote those stages at which E protein activity must be decreased in order for progression to occur. The function at each stage is listed on the right (see text for details). Surface expression of the pre-TCR and TCR during development is shown. HEB is required for expression of the pre-Tα chain gene and the rearrangement and expression of the TCR α locus.

1 Feb 2002

10:17

AR

AR152-11.tex

AR152-11.SGM

LaTeX2e(2001/05/10)

P1: GJC

Annu. Rev. Immunol. 2002.20:301-322. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

E PROTEINS IN LYMPHOID DEVELOPMENT

305

Early B lineage specific transcripts, including Igα, Iµ, RAG1, and Pax5, are lacking in E2A-null mutant bone marrow cells. In contrast, myeloid development is normal. A similar phenotype is observed in mice expressing a transgene for Id1, an inhibitor of E protein activity (40). These mice display the same block in B cell development, and its severity is dependent on the expression level of the transgene. In human bone marrow cells, ectopic expression of Id3 causes an analogous block, which demonstrates the conserved function of E proteins in murine and human lymphopoiesis (41). Two other E protein members, E2-2 and HEB, also have roles in B lineage development (Figure 2a). Mice deficient for HEB or E2-2 generate mature B lineage cells but contain 50% fewer pro-B cells in the fetal liver (42). Interestingly that mice that are transheterozygous for any two of the four E proteins display fewer pro-B cells than mice that are heterozygous for any E protein alone (42). Restoration of E protein expression in an E2A null background allows for B cell development albeit at varying degrees. Expression of both E12 and E47 transgenes in E2A-deficient mice promotes B lineage development better than either transgene alone (43). Furthermore, expression of two (but not one) copies of the HEB gene introduced into the E2A locus results in a partial rescue of B lineage development (44). Mature B cells are generated, but wild-type numbers are not achieved (44). Therefore, the overall timing and dosage of E protein activity may be the key determinant in B lineage progression rather than the activity of a specific E protein. Absence of the EBF transcription factor results in an early B lineage block similar to that observed in E2A-deficient mice (Figure 1) (38). E2A and EBF genetically interact to promote B lineage maturation, as mice that are heterozygous for each of these genes show a more severe B cell phenotype than mice that are heterozygous for either gene alone (45). However, on a per cell basis, the level of VpreB transcripts in double heterozygous mice was shown to be similar to that of wild-type pro-B cells, indicating that the synergy involving E2A and EBF is not at the level of gene expression, but perhaps at the level of cell survival or cellular expansion (45). In addition to E2A and EBF promoting B lineage progression, a key function for Pax5 in establishing B cell commitment was recently demonstrated (46). Pax5deficient B cells arrest during the pro-B cell stage after IgH DJ rearrangement (Figure 1) (39, 47). Thus the absence of Pax5 leads to a block in B cell development at a stage subsequent to that observed in E2A- or EBF-deficient mice. However, pro-B lineage cells derived from Pax5-deficient bone marrow express non-B lineage genes such as pre-Tα, GATA-1, and macrophage colony-stimulating factor (46). Furthermore, B cell progenitors lacking Pax5 maintain the potential to be diverted to other lineages under the appropriate culture conditions (46). Thus, Pax5 expression is essential for commitment to the B cell lineage through its role in suppressing the transcription of genes normally associated with alternative hematopoietic lineages. Since E2A deficiency results in a complete lack of B lineage cells, other strategies have been employed to identify targets of the E2A proteins involved in

1 Feb 2002

10:17

Annu. Rev. Immunol. 2002.20:301-322. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

306

AR

QUONG

AR152-11.tex

¥

AR152-11.SGM

ROMANOW

¥

LaTeX2e(2001/05/10)

P1: GJC

MURRE

initiating the B lineage program. Ectopic expression of E47 in NIH3T3 cells leads to the induction of Ig germline transcription and the expression of other early B cell genes, such as RAG1 and TdT (48). Additionally, overexpression of E12 in 70Z/3 macrophage cells, which carry rearranged IgH and Igκ genes but lack IgM expression, is sufficient to activate the transcription of many B lineage-specific genes including RAG1, λ5, EBF, and Pax5 (49). Interestingly, when EBF is expressed in the 70Z/3 macrophage cell line, transcription of a subset of the E2A-induced genes is upregulated, including Pax5 and λ5 (49). These observations suggest that the E2A proteins regulate the transcription of downstream targets such as Pax5 through its regulation of EBF. Consistent with these observations is the recent finding that the Pax5 promoter contains functional EBF binding sites (45). Moreover, studies suggest that E2A and EBF synergize to induce the expression of common target genes. For example, ectopic expression of both E47 and EBF in a mast cell line, Ba/F3, activates the transcription of the λ5 and VpreB genes (50). Dissection of the promoter of the λ5 gene identified EBF and E47 binding sites, both of which are required for its regulation (50, 51). Taken together, these data suggest that E2A and EBF play dual roles, establishing lineage commitment and promoting B lineage progression, through their ability to activate the expression of genes involved in both processes.

Initiation of V(D)J Recombination and the Pro-B to Pre-B Cell Transition Transition from the pro-B to the pre-B cell stage is dependent on the surface expression of a functionally rearranged IgH chain. Association of IgH chains with surrogate light chains, in conjunction with signaling components Igα and Igβ, constitutes the pre-B cell receptor (BCR). Pre-BCR signaling results in a transient burst of proliferation and allelic exclusion of the IgH chain, followed by initiation of IgL chain gene rearrangement (52–54). B cells from mice that are deficient for either the IgH chain itself or the expression of RAG1 or RAG2 are blocked at the pro-B cell stage (55–57). Similarly, B cell development in mice that are deficient for λ5, a component of the surrogate light chain, fails to progress efficiently to the pre-B cell stage (58, 59). Signaling in pro-B cells mediated through either Igα, Igβ, or an activated form of Ras is sufficient to promote maturation to the pre-B cell stage (60–62). Targets downstream of the pre-BCR remain to be identified, but they are likely to include the E proteins (see below). Recently, the E2A proteins were shown to be required during the IL-7-dependent expansion and survival of pro-B lymphocytes (Figure 2a) (63). In addition to promoting cell survival, E2A proteins are intimately involved with the expression and rearrangement of the IgH and IgL chain genes (Figure 2a). Although it remains to be demonstrated in primary cells, expression of the RAG1 and RAG2 genes has been shown to be regulated by the E2A proteins (45, 49). E2A proteins have the ability to promote IgH and IgL chain gene rearrangement in various cell types. Ectopic expression of E47 in the pre-T cell line, 2017, causes the induction of

1 Feb 2002

10:17

AR

AR152-11.tex

AR152-11.SGM

LaTeX2e(2001/05/10)

P1: GJC

Annu. Rev. Immunol. 2002.20:301-322. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

E PROTEINS IN LYMPHOID DEVELOPMENT

307

IgH D to J rearrangement, accompanied by an increase in the level of Iµ germline transcripts (64). Furthermore, expression of E47, in addition to RAG1 and RAG2, in an embryonic kidney cell line promotes both IgH DJ and Igκ VJ rearrangements (65). Rearrangements of the Igκ locus predominantly used Vκ1-family segments, which contain a conserved E box in their promoter regions. Ig rearrangements in this system were similar to those seen in endogenous B cells: distinct V segments were utilized and the coding joints were diverse and often contained deletions (65). These data suggest that E2A proteins play a key role in the regulation of both IgH and Igκ recombination. Since, in vivo, these rearrangements typically occur sequentially rather than concurrently, it is likely that additional transcriptional regulators act together with the E2A proteins to promote the ordered rearrangement of the IgH and IgL genes. One such candidate is EBF, which has been shown to promote IgH DJ but not Vκ-J joining (65, 65a). Additionally, it is possible that E protein levels are differentially regulated in pro-B versus pre-B cells. It will be important to examine carefully the relative levels of the individual E proteins in pro-B and pre-B cells and to determine whether E protein activity is regulated by signals mediated by the pre-BCR similar to that described for TCR-mediated signaling (see below).

Antigen Dependent Maturation of B Lineage Cells Antigen engagement of the mature B cell receptor initiates B cell activation characterized by cell cycle entry and upregulation of specific surface markers. Activated B cells receiving co-stimulatory T cell derived signals differentiate to become low affinity plasma cells or migrate to form germinal centers. There they undergo clonal expansion, affinity maturation and immunoglobulin isotype switching, resulting in high-affinity antibody-secreting effector cells and memory B cells (66–69). In the peripheral lymphoid organs, the E2A proteins are present at low but detectable levels in resting na¨ıve B cells (70). In contrast, E2A protein expression is high in activated, mature B cells and in cells present in the dark zone of germinal centers (70, 71). In vitro activation of B lymphocytes through co-stimulation with T cells and antigen or through various mitogenic stimuli results in higher levels of E2A proteins and DNA binding activity (70). Thus, induction of E2A activity appears to be a common feature during B cell activation. Id3 transcripts have been shown to be rapidly induced upon BCR engagement, suggesting a role in promoting cell cycle entry (72, 73). Consistent with this model, Id3-deficient mice display a proliferation defect in response to anti-IgM stimulation (72). The increase in Id3 levels is transient and occurs prior to the induction of E2A activity (70, 73). Whether Id3 regulates subsequent events during antigen-induced differentiation remains to be elucidated, but it is plausible that Id3 functions, at least in part, by regulating the activity of the E proteins. In contrast to pro-B lymphocyte expansion, loss of E2A activity in peripheral activated B cells does not interfere with proliferation and survival (70). In addition, the induction and regulation of early and late activation markers, including CD69,

1 Feb 2002

10:17

Annu. Rev. Immunol. 2002.20:301-322. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

308

AR

QUONG

AR152-11.tex

¥

AR152-11.SGM

ROMANOW

¥

LaTeX2e(2001/05/10)

P1: GJC

MURRE

CD44, IgD, and IgM, is not impaired. However, activated B lymphocytes do not undergo isotype switching in response to appropriate stimuli in the absence of E2A activity (Figure 2a) (70). Similarly, ectopic expression of Id1 in two human B cell lines undergoing isotype switching significantly interferes with IgA surface expression (71). Activated primary B lymphocytes lacking E2A activity are blocked in class switching at the level of genomic recombination (70). Since the appropriate switch region germline transcripts are present, it is unlikely that the primary defect in switching is caused by a lack of chromatin accessibility (70, 71). Rather, we propose that the E2A proteins are required for the expression of components of the switch recombinase machinery.

THE ROLES OF E PROTEINS IN T LYMPHOCYTE DEVELOPMENT Thymocyte Maturation T cell development can be characterized based on the rearrangement status of the TCR loci and the expression of the CD4 and CD8 co-receptors (Figure 3). The earliest T cell progenitors in the thymus are present within the CD4 and

Figure 3 Schematic representation of murine T lymphocyte development. The developmental stages are shown together with their characteristic cell surface markers. Transcription factors listed above the arrows denote the approximate developmental stages at which they have been shown to function. The progression of β selection deficient DP cells in the absence of E2A to the DP stage is denoted by the long arrow from the DN3 to DP stage (see text). The appropriate stages when TCR rearrangements are incurred are indicated by the horizontal lines. Vertical dashed lines delineate pre-TCR and TCR mediated checkpoints required for developmental progression.

1 Feb 2002

10:17

AR

AR152-11.tex

AR152-11.SGM

LaTeX2e(2001/05/10)

P1: GJC

Annu. Rev. Immunol. 2002.20:301-322. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

E PROTEINS IN LYMPHOID DEVELOPMENT

309

CD8 double negative (DN) population. These cells are characterized as being CD44+ and CD25− and still have the ability to mature into NK and T lineage cells. Commitment to the T lineage is largely associated with the activation of CD25 expression and is followed by initiation of gene rearrangements at the TCR γ , δ, and β loci. The formation of a functional TCR β gene product allows for the expression of the pre-TCR complex, which includes the TCR β chain as well as the pre-Tα protein and the CD3 assembly of signaling molecules. Signaling mediated by the pre-TCR complex results in developmental progression, also referred to as β selection. This transition is characterized by inhibition of gene rearrangement, initiation of cellular expansion, and the maturation into CD4 and CD8 double positive (DP) thymocytes. DP cells then exit the cell cycle and begin TCR α gene rearrangement. The expression of an αβ TCR allows DP cells to undergo major histocompatibility complex (MHC)-mediated positive or negative selection. Positively selected DP thymocytes downregulate either CD4 or CD8 expression to become single positive (SP) mature T lineage cells (74–77).

The T versus NK Lineage Decision NK and T lineage cells develop from a bipotent T/NK precursor cell (78–80). Recent evidence has indicated a role for HLH proteins in both NK and T lineage commitment (Figure 3). Mice that are deficient for Id2 display a block in NK cell development at the T/NK precursor cell stage (81). In contrast, αβ T cell development is normal. E2A-deficient mice display a complimentary phenotype characterized by a partial block in T cell development and normal NK cell development (23). Similarly, forced expression of Id3, another E2A antagonist, in precursor T/NK cells or in committed progenitor T cells, blocks T lineage progression but not NK cell development (82, 83). Taken together, these observations indicate a key role for HLH proteins at this stage and demonstrate that the ratio of E2A and Id2 likely regulates the NK versus T cell lineage decision.

Expression of E2A during T Lineage Development The expression of E47 in γ δ and αβ T lineage cells has been examined in great detail. The majority of CD44+ CD25– (DN1) thymocytes represents cells that have not yet committed to the T cell lineage and express little or no E47 (84). The transition to the CD44+ CD25+ (DN2) stage is closely correlated with a bias towards T cell development. Interestingly, E47 expression is induced at the DN2 stage. In addition, γ δ T cells, which diverge from the αβ T lineage, express elevated levels of E47 (I. Engel, C. Murre, unpublished observations). E47 levels remain high throughout DN thymocyte maturation; however, upon pre-TCR mediated signaling, E47 DNA-binding activity is significantly downregulated (84). As compared to DN thymocytes, DP cells express lower levels of E47 proteins, which are decreased further during the transition from DP to SP cells. These

1 Feb 2002

10:17

Annu. Rev. Immunol. 2002.20:301-322. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

310

AR

QUONG

AR152-11.tex

¥

AR152-11.SGM

ROMANOW

¥

LaTeX2e(2001/05/10)

P1: GJC

MURRE

Figure 4 Model of T lymphocyte progression and E protein expression. Relative E47 protein levels in DN, DP and SP thymocytes (as described in text) are indicated as the plateaus. Progression from one stage to the next is initiated by pre-TCR/TCR signaling (vertical arrows) resulting in increases in Id3 transcripts and decreased E protein DNA binding activity (as described in text). T lineage developmental progression is represented by the x axis. The rate of decrease of E2A at each transition is conjectured and remains to be determined.

observations indicate that a gradient of E47 expression is present during thymocyte development: E47 levels are highest in DN cells and lowest in SP cells (Figure 4) (84).

Early T Lineage Development Consistent with the expression pattern of E47 in thymocytes, loss of E2A affects both the αβ and γ δ T cell lineages (Figures 2b and 3). In E2A-deficient mice, not only are γ δ T cell numbers reduced, but the temporal regulation of TCR γ and δ recombination is disrupted (85). E2A is required to repress fetal-specific γ and δ rearrangements in adult thymocytes, while promoting the rearrangement of the appropriate adult V segments (85). Furthermore, E2A and HEB, in conjunction

1 Feb 2002

10:17

AR

AR152-11.tex

AR152-11.SGM

LaTeX2e(2001/05/10)

P1: GJC

Annu. Rev. Immunol. 2002.20:301-322. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

E PROTEINS IN LYMPHOID DEVELOPMENT

311

with RAG1 and RAG2, have the ability to induce γ δ TCR rearrangements in an embryonic kidney cell line (86, 87). αβ T cell development in E2A-deficient mice is also affected. Thymocyte cell numbers are decreased five- to ten-fold, and thymocyte development is partially blocked at the DN1 to DN2 transition (23). During the DN stages of αβ T cell development, rearrangement of the TCR β locus occurs. In addition to the induction of the RAG recombination machinery by E2A, evidence suggests that ablation of E protein activity, through expression of a dominant-negative form of HEB or an Id1 transgene, results in a decrease in Vβ to DJβ rearrangements (88, 89). Taken together, these data suggest that E protein activity is necessary for efficient progression toward the T cell lineage as well as directing early events such as V(D)J recombination in T cell development.

The DN to DP Transition Upon successful rearrangement of the TCR β locus, the TCR β chain pairs with pre-Tα and the CD3 subunits to form the pre-TCR complex. Signaling through this complex is essential for thymocytes to mature from the DN to the DP stage, a process referred to as β selection (90). HEB is necessary for the expression of the pre-Tα gene as demonstrated in HEB-deficient and SCL/LMO transgenic mice (91). Similarly, transduction of human pre-T cells with Id3 results in the suppression of pre-Tα transcription and an analogous block (83). E2A also functions to regulate β selection (Figure 2b). Genetic data have recently indicated that E47 deficiency is sufficient to promote the maturation of DN cells to the DP cell stage in the absence of a functional TCR β chain (84). This implies that pre-TCR signaling functions to promote progression, in part, by inhibiting E2A activity. Consistent with this model, E47 expression remains high during DN thymocyte maturation; however, upon pre-TCR-mediated signaling, E47 DNA binding activity is downregulated. In conjunction, Id3 RNA levels are increased upon pre-TCR-mediated signaling in DN thymocytes through a pathway dependent upon the ERK MAPK cascade (84). Taken together, these observations provide both genetic and biochemical evidence that E protein activity is downregulated by pre-TCR signaling to promote β selection (Figure 4). Once the DN T cells pass β selection, CD8 expression is upregulated, resulting in intermediate single positive (ISP) cells, followed by progression to the DP stage. Thymii of HEB-deficient mice display an increased proportion of ISP cells due to a block in the transition to the DP stage (92). How HEB functions during this transition remains to be clarified; however, regulation of the CD4 gene by HEB may play a role (9).

Thymocyte Selection Upon reaching the DP cell stage, thymocytes initiate TCRα rearrangement. Once a functional αβ TCR has formed, DP cells interact with peptide-MHC complexes

1 Feb 2002

10:17

Annu. Rev. Immunol. 2002.20:301-322. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

312

AR

QUONG

AR152-11.tex

¥

AR152-11.SGM

ROMANOW

¥

LaTeX2e(2001/05/10)

P1: GJC

MURRE

to undergo selection. DP cells receiving the appropriate signal mediated by the TCR are positively selected. In contrast, cells expressing a TCR with high affinity for self-peptide-MHC complexes undergo negative selection. Failure to receive a TCR-mediated signal results in death by neglect (93–95). E2A-deficient mice have decreased proportions of DP cells and elevated percentages of SP T cells with a particular skewing toward the CD8 SP population (23). The decrease in DP cells is due in part to reduced viability (96). In fact, massive apoptosis is observed in thymocytes from Id1 transgenic mice, attributable to the fact that both E2A and HEB are inhibited (88). In addition, enhancement of both MHC class I- and class II-mediated positive selection is observed in E2Adeficient mice (96). In particular, E47-deficient, H-Y TCR transgenic female mice have increased percentages of CD8 SP cells in both the thymus and peripheral tissues. The acceleration of selection to the CD8 SP lineage in the thymus is evident in mice heterozygous for E47 (96). Thus, selection to the CD8 lineage appears sensitive to E protein dosage. Recently, studies analyzing E47 deficiency in mice with blocks in TCR β rearrangement have demonstrated that not only is progression to the DP stage restored, but development of a small number of CD8 SP thymocytes can be detected (84). It is interesting that CD4 SP cells are consistently absent, which indicates that a loss of E2A is sufficient to drive CD8 SP positive selection, whereas additional signals appear to be required for maturation to the CD4 lineage. Although it remains to be determined if the effects in positive selection are T cell intrinsic, these data suggest an important role for E2A in attenuating thymocyte selection (Figure 2b). Id3 has been shown to interact genetically with E2A in the thymus. Mice deficient for Id3 display a thymic phenotype complimentary to that of E2A-deficient mice (33, 96). Moreover, mice lacking both E2A and Id3 display a relieved thymic phenotype with respect to either deficiency alone (33). Total thymic cellularity in Id3-deficient mice is normal; however, decreased percentages and numbers of CD4 SP cells are observed in the thymus, and decreases of both CD4 and CD8 SP cells are observed in the spleen (33). When analyzed in the AND and H-Y TCR transgenic backgrounds, both MHC class II- and MHC class I-restricted positive selection, respectively, are inhibited in Id3-deficient mice. In particular, CD4 positive selection in AND TCR transgenic mice is almost completely blocked in the absence of Id3. Furthermore, the effect of Id3 on positive selection is intrinsic to the T cell lineage (Figure 3) (33). A role for Id3 in the negative selection process has also been demonstrated (33). Thymocytes from male H-Y TCR transgenic mice recognize the male specific antigen as self and are deleted. Loss of Id3 in these mice results in increased thymic cellularity, considerably more DP thymocytes and increases in CD4 SP cells in the spleen. In contrast, superantigen-mediated deletion appears normal in Id3−/− mice in the MHC class II I-Ed BALB/c background (33). Preliminary results indicate that the defect in negative selection observed in H-Y TCR transgenic male Id3-deficient mice is not intrinsic to the T cell lineage and remains to be further examined (R. Rivera, C. Murre, unpublished results).

1 Feb 2002

10:17

AR

AR152-11.tex

AR152-11.SGM

LaTeX2e(2001/05/10)

P1: GJC

E PROTEINS IN LYMPHOID DEVELOPMENT

313

SIGNALING PATHWAYS REGULATING E PROTEIN FUNCTION

Annu. Rev. Immunol. 2002.20:301-322. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

Regulation of E Protein Activity by the Ras/ERK MAP Kinase Cascade Thymocytes expressing a dominant-negative form of MAPKK1 (MEK1) fail to progress from the DN to the DP stage in fetal thymic organ culture (97). Furthermore, mice lacking Ras activity or the downstream effectors MEK1 or ERK1 display defective thymocyte maturation at the DP to SP transition reminiscent of the block in Id3-deficient mice (33, 98, 99). As mentioned above, E proteins appear to be functioning at these checkpoints (Figure 4) (84, 100). From these observations, it was postulated that TCR-mediated signals promote developmental progression through the modulation of E protein DNA binding and/or transcriptional activation (Figure 4). In fact, recent observations have demonstrated that signals mediated by both the pre-TCR and the αβ TCR act to downmodulate E2A/HEB DNA binding activity (84, 100). This decrease is caused, in part, by a decrease in E2A protein levels during thymocyte maturation. Additionally, Id3 levels are increased in response to CD3 mediated signals at both the DN and DP stages (84, 100). Whereas it remains to be determined how varying levels of E47 protein are established, the regulation of Id3 has been characterized in great detail. Mice carrying a dominant negative form of Ras do not induce Id3 transcription in response to TCR crosslinking (100). Inhibition of the downstream target c-Raf, using the pharmacological drug PD98059, blocks the induction of Id3 transcripts in a dosedependent manner. Additional evidence indicates that the transcription of Id3 is regulated, at least in part, by the ERK MAPK cascade and is mediated by the immediate early transcription factor, EGR-1 (100). The dose-dependent induction of Id3 transcripts in response to ERK MAPK signaling raises the interesting possibility that the magnitude of the signal determines the level of E protein activity (100). Although more evidence is required, the observed bias toward the CD8 lineage in the absence of E2A suggests a role for E proteins in the CD4 versus CD8 lineage decision (23). Thus, it is conceivable that a strong signal mediated by the ERK MAP kinase pathway leads to high levels of Id3, favoring CD4 lineage commitment, whereas a weak signal would only modestly affect Id3 levels, promoting CD8 development (101). It will be particularly interesting to modulate the relative levels of E2A, HEB, and Id3 and to examine how these ratios affect cell fate.

Regulation of Id Expression by TGF-β and c-myc In addition to its regulation by the Ras-ERK MAP kinase pathway, Id gene expression has also been shown to be regulated by members of the TGF-β family (63). For example, bone morphogenetic protein 4 (BMP4) has been shown to activate both Id2 and Id3 transcription in embryonic stem cells (102). Furthermore, in

1 Feb 2002

10:17

Annu. Rev. Immunol. 2002.20:301-322. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

314

AR

QUONG

AR152-11.tex

¥

AR152-11.SGM

ROMANOW

¥

LaTeX2e(2001/05/10)

P1: GJC

MURRE

primary B lymphocyte progenitors (BLP) TGF-β induces cell cycle growth arrest through the transient induction of Id3 transcription (63). Ectopic expression of Id3 alone is sufficient for growth arrest, and Id3-deficient BLPs respond more slowly to TGF-β signaling. Unlike in DP thymocytes, the mechanism of Id3 induction by TGF-β appears to involve the activation of Smad transcription factors and not the Ras-ERK module (63). Thus, distinct signaling pathways are utilized to modulate E protein activity by targeting Id gene expression. Recent observations have suggested a role for c-myc in controlling Id2 gene transcription (103). Enforced expression of N- or c-myc induced Id2 expression in a number of cell lines. It was postulated that myc induction of Id2 leads to the inactivation of Rb, thereby driving cellular proliferation (103). However, we note that it is plausible that increases in Id2 levels promote cellular transformation through direct inhibition of E protein activity (discussed below).

E2A and Notch-Mediated Signaling Recent data has implicated Notch-mediated signaling in the decision between B and T lineage determination (Figures 1 and 3). Mice conditionally deficient for Notch1 are blocked at the progenitor T cell stage (104). Conversely, mice that are transgenic for a constitutively activated form of Notch share many phenotypes in common with E2A-deficient mice: (a) B cell development is blocked prior to the onset of Ig gene rearrangement; (b) T cell lymphomas develop with similar kinetics; and (c) abnormalities are observed in both the αβ and γ δ T cell lineages (105, 106). These observations have led to the suggestion that E2A activity is regulated in part by the Notch signaling pathway. In support of this, an activated form of Notch has been shown to repress E2A activity in transient transfection assays (107). Further in vivo studies will be required to determine if Notch and E2A act in a linear pathway or, alternatively, act in parallel.

E PROTEIN ACTIVITY, LYMPHOCYTE SURVIVAL, PROLIFERATION AND MALIGNANCY The E2A proteins have been implicated in cellular proliferation and apoptosis. Several studies have shown that both high and low levels of E2A activity promote rapid cell death, which suggest that lymphocytes can only survive within a limited range of E2A activity (88, 108, 109). Mice deficient for E2A rapidly develop highly malignant T cell lymphoma (23, 110). Similarly, ectopic expression of Id1 or Id2 in transgenic mice promotes lymphomagenesis (88, 111). Restoration of E2A activity into cells derived from E2A-deficient lymphomas leads to apoptosis (108). Likewise, ectopic expression of E47 in human T cell acute lymphoblastic leukemia (T-ALL) cell lines results in a block in cell cycle progression and activated programmed cell death (109). These observations have led to the proposal that E2A functions as a tumor suppressor and its inactivation is a key factor in the development of human T-ALL (108, 109).

1 Feb 2002

10:17

AR

AR152-11.tex

AR152-11.SGM

LaTeX2e(2001/05/10)

P1: GJC

Annu. Rev. Immunol. 2002.20:301-322. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

E PROTEINS IN LYMPHOID DEVELOPMENT

315

E2A is also involved in a significant percentage of childhood pro-B and preB cell leukemias resulting from chromosomal translocations. In pre-B cell ALL resulting from a t(1;19) translocation, the 50 portion of the E2A gene is fused to the homeodomain-containing region of the pbx1 gene (112). This translocation event results in the expression of a chimeric E2A-pbx fusion protein. In pro-B cell leukemias containing a t(17;19) translocation, the E2A gene is fused to HLF, a gene encoding a bZIP protein (113). The exact mechanism by which these translocations cause leukemia has not yet been elucidated. It has been postulated that the targetting of the transactivation domains of E2A to alternative loci by a heterologous DNAbinding domain results in the aberrant expression of genes associated with cell growth and/or survival ultimately leading to cellular transformation. However, these translocation events also effectively reduce the copy number and expression of the E2A gene by half. Although it remains to be proven, reduction of E2A activity may, in fact, be an essential event in the transformation of B cell progenitors in the same way that E2A functions as a tumor suppressor in the T cell lineage. Various signaling molecules, including p56Lck, TPL2/Cot, MEK1, and most notably, activated forms of Ras, have been implicated in the development of T cell lymphomas (114–116). For example, overexpression of p56Lck in developing thymocytes results in the rapid development of T cell tumors (114). The oncogenic protein TPL2/Cot is a serine kinase with the ability to phosphorylate MEK1 (116–118). It is conceivable that its expression inappropriately activates Id gene expression in T lineage cells, ultimately resulting in lymphoma. Interestingly, activated forms of Ras and high levels of Id gene expression have been detected in pancreatic cancers (119–121). It will be important to determine if malignancies that possess activated Ras have become transformed through the induction of Id gene expression.

CONCLUSION The development of B and T lymphocytes is dependent on the temporal activation of lineage-specific genes, productive antigen receptor rearrangements, as well as the coordination of survival and proliferation with developmental progression. E proteins have now been shown to function in all of these processes (Figure 2). A number of target genes of E proteins have already been identified in lymphocytes. The E proteins regulate the expression of the RAG genes and invariant receptor chains λ5, VpreB and pre-Tα, and promote Ig and TCR gene rearrangements. Thus, the target genes of these proteins perform similar functions in both B and T lymphocytes. However, it is in the T cell lineage that the role of E proteins as downstream effectors of signaling has been elucidated. The E proteins have been demonstrated to be downstream of TCR signaling via the ERK MAP kinase pathway, acting to regulate lineage progression between the DN to DP and DP to SP cell stages (Figure 4). Furthermore, E proteins function to inhibit both proliferation and developmental progression in the absence of rearrangements. In B cells, a role for E proteins in pre-BCR signaling has not yet been demonstrated.

1 Feb 2002

10:17

Annu. Rev. Immunol. 2002.20:301-322. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

316

AR

QUONG

AR152-11.tex

¥

AR152-11.SGM

ROMANOW

¥

LaTeX2e(2001/05/10)

P1: GJC

MURRE

However, we note that in mature B lineage cells, E2A protein levels are regulated by signals emanating from the BCR. Given that Ras signaling is essential at the pro-B to pre-B cell transition, it is conceivable that the E proteins are regulated similarly to promote B lineage progression. A key question that requires further analysis is how these and other target genes of E proteins are regulated in a tissue- and temporal-specific fashion. Ultimately, this will require a precise and detailed analysis of the regulatory elements that control the transcription of these genes. A comprehensive understanding of how HLH proteins function in lymphocyte development will have to include additional studies of the mechanisms by which E proteins are regulated. For example, it will be important to determine how differing levels of E47 expression are formed during thymocyte development, whether similar gradients are formed during B lineage differentiation and whether they are regulated by antigen receptor mediated signaling. In addition, modifications to the E proteins and their effects on transcriptional activity will need to be further investigated (122, 123). Finally, the identification of novel targets involved in promoting survival and proliferation will give insights into how the E proteins coordinate cell survival and developmental progression. ACKNOWLEDGMENTS We would like to thank Dr. Eben Massari for critical reading of the manuscript and members of the Murre Lab for helpful discussions. This work was supported by grants from the National Institutes of Health (NIH). W.J.R is supported by a Developmental Biology NIH Training Grant. Visit the Annual Reviews home page at www.annualreviews.org

LITERATURE CITED 1. Kondo M, Weissman IL, Akashi K. 1997. Identification of clonogenic common lymphoid progenitors in mouse bone marrow. Cell 91:661–72 2. Ephrussi A, Church GM, Tonegawa S, Gilbert W. 1985. B lineage-specific interactions of an immunoglobulin enhancer with cellular factors in vivo. Science 227:134–40 3. Henthorn P, Kiledjian M, Kadesch T. 1990. Two distinct transcription factors that bind the immunoglobulin enhancer µE5/κE2 motif. Science 247:467–70 4. Murre C, McCaw PS, Baltimore D. 1989. A new DNA binding and dimerization motif in immunoglobulin enhancer bind-

ing, daughterless, MyoD, and myc proteins. Cell 56:777–83 5. Ho IC, Yang LH, Morle G, Leiden JM. 1989. A T-cell-specific transcriptional enhancer element 30 of Cα in the human Tcell receptor α locus. Proc. Natl. Acad. Sci. USA 86:6714–18 6. Reizis B, Leder P. 1999. Expression of the mouse pre-T cell receptor α gene is controlled by an upstream region containing a transcriptional enhancer. J. Exp. Med. 189:1669–78 7. Ernst P, Smale ST. 1995. Combinatorial regulation of transcription II: the immunoglobulin µ heavy chain gene. Immunity 2:427–38

1 Feb 2002

10:17

AR

AR152-11.tex

AR152-11.SGM

LaTeX2e(2001/05/10)

P1: GJC

Annu. Rev. Immunol. 2002.20:301-322. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

E PROTEINS IN LYMPHOID DEVELOPMENT 8. Staudt LM, Lenardo MJ. 1991. Immunoglobulin gene transcription. Annu. Rev. Immunol. 9:373–98 9. Sawada S, Littman DR. 1993. A heterodimer of HEB and an E12-related protein interacts with the CD4 enhancer and regulates its activity in T-cell lines. Mol. Cell. Biol. 13:5620–28 10. Kee BL, Murre C. 2001. Transcription factor regulation of B lineage commitment. Curr. Opin. Immunol. 13:180–85 11. Massari ME, Murre C. 2000. Helix-loophelix proteins: regulators of transcription in eucaryotic organisms. Mol. Cell. Biol. 20:429–40 12. Murre C, McCaw PS, Vaessin H, Caudy M, Jan LY, Jan YN, Cabrera CV, Buskin JN, Hauschka SD, Lassar AB, Weintraub BC, Baltimore D. 1989. Interactions between heterologous helix-loop-helix proteins generate complexes that bind specifically to a common DNA sequence. Cell 58:537–44 13. Quong MW, Massari ME, Zwart R, Murre C. 1993. A new transcriptionalactivation motif restricted to a class of helix-loop-helix proteins is functionally conserved in both yeast and mammalian cells. Mol. Cell. Biol. 13:792–800 14. Aronheim A, Shiran R, Rosen A, Walker MD. 1993. The E2A gene product contains two separable and functionally distinct transcription activation domains. Proc. Natl. Acad. Sci. USA 90:8063–67 15. Massari ME, Jennings PA, Murre C. 1996. The AD1 transactivation domain of E2A contains a highly conserved helix which is required for its activity in both Saccharomyces cerevisiae and mammalian cells. Mol. Cell. Biol. 16:121–29 16. Massari ME, Grant PA, Pray-Grant MG, Berger SL, Workman JL, Murre C. 1999. A conserved motif present in a class of helix-loop-helix proteins activates transcription by direct recruitment of the SAGA complex. Mol. Cell 4:63–73 17. Sun XH, Baltimore D. 1991. An inhibitory domain of E12 transcription factor

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

317

prevents DNA binding in E12 homodimers but not in E12 heterodimers. Cell 64:459–70 Hu JS, Olson EN, Kingston RE. 1992. HEB, a helix-loop-helix protein related to E2A and ITF2 that can modulate the DNA-binding ability of myogenic regulatory factors. Mol. Cell. Biol. 12:1031–42 Murre C, Baltimore D. 1992. The helixloop-helix motif: structure and function. In Transcriptional Regulation, ed. SL McKnight, KR Yamamoto, pp. 861–79. Plainview, NY: Cold Spring Harbor Lab. Murre C, Voronova A, Baltimore D. 1991. B-cell- and myocyte-specific E2box-binding factors contain E12/E47-like subunits. Mol. Cell. Biol. 11:1156–60 Bain G, Gruenwald S, Murre C. 1993. E2A and E2-2 are subunits of B-cellspecific E2-box DNA-binding proteins. Mol. Cell. Biol. 13:3522–29 Shen CP, Kadesch T. 1995. B-cellspecific DNA binding by an E47 homodimer. Mol. Cell. Biol. 15:4518–24 Bain G, Engel I, Robanus Maandag EC, te Riele HPJ, Voland JR, Sharp LL, Chun J, Huey B, Pinkel D, Murre C. 1997. E2A deficiency leads to abnormalities in αβ T-cell development and to rapid development of T-cell lymphomas. Mol. Cell. Biol. 17:4782–91 Rivera RR, Murre C. 2001. The regulation and function of the Id proteins in lymphocyte development. Oncogene. In press Benezra R, Davis RL, Lockson D, Turner DL, Weintraub H. 1990. The protein Id-a negative regulator of helix-loop-helix DNA binding proteins. Cell 61:49–59 Christy B, Sanders L, Lau L, Copeland N, Jenkins N, Nathans D. 1991. An Idrelated helix-loop-helix protein encoded by a growth-factor inducible gene. Proc. Natl. Acad. Sci. USA 88:1815–19 Deed RW, Bianchi SM, Atherton GT, Johnston D, Santibanez-Koref M, Murphy JJ, Norton JD. 1992. An immediate early human gene encodes an Id-like helix-loop-helix protein and is regulated

1 Feb 2002

10:17

318

28.

Annu. Rev. Immunol. 2002.20:301-322. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

29.

30.

31.

32.

33.

34.

35.

36.

AR

QUONG

AR152-11.tex

¥

AR152-11.SGM

ROMANOW

¥

LaTeX2e(2001/05/10)

P1: GJC

MURRE

by protein kinase C activation in diverse cell types. Oncogene 8:599–607 Ellis HM, Spann DR, Posakony JW. 1990. Extramacrochaetae, a negative regulator of sensory organ development in Drosophila, defines a new class of helixloop-helix proteins. Cell 61:27–38 Garrell J, Modolell J. 1990. The Drosophila extramacrochaetae locus, an antagonist of proneural genes that, like these genes, encodes a helix-loop-helix protein. Cell 61:39–48 Riechmann V, Cruchten Iv, Sablitzky F. 1994. The expression pattern of Id4, a novel dominant negative helix-loop-helix protein, is distinct from Id1, Id2 and Id3. Nucleic Acids Res. 22:749–55 Sun XH, Copeland NG, Jenkins NA, Baltimore D. 1991. Id proteins Id1 and Id2 selectively inhibit DNA binding by one class of helix-loop-helix proteins. Mol. Cell. Biol. 11:5603–11 Kee BL, Quong MW, Murre C. 2000. E2A proteins: essential regulators at multiple stages of B-cell development. Immunol. Rev. 175:138–49 Rivera RR, Johns CP, Quan J, Johnson RS, Murre C. 2000. Thymocyte selection is regulated by the helix-loop-helix inhibitor protein, Id3. Immunity 12:17–26 Yu W, Misulovin Z, Suh H, Hardy RR, Jankovic M, Yannoutsos N, Nussenzweig MC. 1999. Coordinate regulation of RAG1 and RAG2 by cell type-specific DNA elements 50 of RAG2. Science 285:1080–84 Nagaoka H, Yu W, Nussenzweig MC. 2000. Regulation of RAG expression in developing lymphocytes. Curr. Opin. Immunol. 12:187–90 Bain G, Maandag EC, Izon DJ, Amsen D, Kruisbeek AM, Weintraub BC, Krop I, Schlissel MS, Feeney AJ, van Roon M, van der Valk M, te Riele HPJ, Berns A, Murre C. 1994. E2A proteins are required for proper B cell development and initiation of immunoglobulin gene rearrangements. Cell 79:885–92

37. Zhuang Y, Soriano P, Weintraub H. 1994. The helix-loop-helix gene E2A is required for B cell formation. Cell 79:875–84 38. Lin H, Grosschedl R. 1995. Failure of B-cell differentiation in mice lacking the transcription factor EBF. Nature 376:263– 67 39. Nutt SL, Urbanek P, Rolink A, Busslinger M. 1997. Essential functions of Pax5 (BSAP) in pro-B cell development: difference between fetal and adult B lymphopoiesis and reduced V-to-DJ recombination at the IgH locus. Genes Dev. 11: 476–91 40. Sun XH. 1994. Constitutive expression of the Id1 gene impairs mouse B cell development. Cell 79:893–900 41. Jaleco AC, Stegmann AP, Heemskerk MH, Couwenberg F, Bakker AQ, Weijer K, Spits H. 1999. Genetic modification of human B-cell development: B-cell development is inhibited by the dominant negative helix loop helix factor Id3. Blood 94:2637–46 42. Zhuang Y, Cheng P, Weintraub H. 1996. B-lymphocyte development is regulated by the combined dosage of three basic helix-loop-helix genes, E2A, E2-2, and HEB. Mol. Cell. Biol. 16:2898–905 43. Bain G, Robanus Maandag EC, te Riele HPJ, Feeney AJ, Sheehy A, Schlissel M, Shinton SA, Hardy RR, Murre C. 1997. Both E12 and E47 allow commitment to the B cell lineage. Immunity 6:145–54 44. Zhuang Y, Barndt RJ, Pan L, Kelley R, Dai M. 1998. Functional replacement of the mouse E2A gene with a human HEB cDNA. Mol. Cell. Biol. 18:3340–49 45. O’Riordan M, Grosschedl R. 1999. Coordinate regulation of B cell differentiation by the transcription factors EBF and E2A. Immunity 11:21–31 46. Nutt SL, Heavey B, Rolink AG, Busslinger M. 1999. Commitment to the B-lymphoid lineage depends on the transcription factor Pax5. Nature 401:556–62 47. Urbanek P, Wang ZQ, Fetka I, Wagner EF, Busslinger M. 1994. Complete block

1 Feb 2002

10:17

AR

AR152-11.tex

AR152-11.SGM

LaTeX2e(2001/05/10)

P1: GJC

E PROTEINS IN LYMPHOID DEVELOPMENT

48.

Annu. Rev. Immunol. 2002.20:301-322. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

49.

50.

51.

52.

53.

54.

55.

56.

57.

of early B cell differentiation and altered patterning of the posterior midbrain in mice lacking Pax5/BSAP. Cell 79:901–12 Choi JK, Shen CP, Radomska HS, Eckhardt LA, Kadesch T. 1996. E47 activates the Ig-heavy chain and TdT loci in non-B cells. EMBO J. 15:5014–21 Kee BL, Murre C. 1998. Induction of early B cell factor (EBF) and multiple B lineage genes by the basic helix-loophelix transcription factor E12. J. Exp. Med. 188:699–713 Sigvardsson M, O’Riordan M, Grosschedl R. 1997. EBF and E47 collaborate to induce expression of the endogenous immunoglobulin surrogate light chain genes. Immunity 7:25–36 Sigvardsson M. 2000. Overlapping expression of early B-cell factor and basic helix-loop-helix proteins as a mechanism to dictate B-lineage-specific activity of the lambda5 promoter. Mol. Cell. Biol. 20:3640–54 L¨offert D, Schaal S, Ehlich A, Hardy RR, Zou YR, M¨uller W, Rajewsky K. 1994. Early B-cell development in the mouse: insights from mutations introduced by gene targeting. Immunol. Rev. 137:135– 53 Yankee TM, Clark EA. 2000. Signaling through the B cell antigen receptor in developing B cells. Rev. Immunogenet. 2:185–203 Hardy RR, Hayakawa K. 2001. B cell development pathways. Annu. Rev. Immunol. 19:595–621 Kitamura D, Roes J, Kuhn R, Rajewsky K. 1991. A B cell-deficient mouse by targeted disruption of the membrane exon of the immunoglobulin µ chain gene. Nature 350:423–26 Mombaerts P, Iacomini J, Johnson RS, Herrup K, Tonegawa S, Papaioannou VE. 1992. RAG-1-deficient mice have no mature B and T lymphocytes. Cell 68:869–77 Shinkai Y, Rathbun G, Lam KP, Oltz EM, Stewart V, Mendelsohn M, Charron J, Datta M, Young F, Stall AM, Alt FW.

58.

59.

60.

61.

62.

63.

64.

65.

65a.

319

1992. RAG-2-deficient mice lack mature lymphocytes owing to inability to initiate V(D)J rearrangement. Cell 68:855–67 Kitamura D, Kudo A, Schaal S, Muller W, Melchers F, Rajewsky K. 1992. A critical role of lambda 5 protein in B cell development. Cell 69:823–31 Papavasiliou F, Jankovic M, Nussenzweig MC. 1996. Surrogate or conventional light chains are required for membrane immunoglobulin mu to activate the precursor B cell transition. J. Exp. Med. 184:2025–30 Teh YM, Neuberger MS. 1997. The immunoglobulin (Ig) α and Ig β cytoplasmic domains are independently sufficient to signal B cell maturation and activation in transgenic mice. J. Exp. Med. 185:1753– 58 Shaw AC, Swat W, Ferrini R, Davidson L, Alt FW. 1999. Activated Ras signals developmental progression of recombinase-activating gene (RAG)-deficient pro-B lymphocytes. J. Exp. Med. 189: 123–29 Shaw AC, Swat W, Davidson L, Alt FW. 1999. Induction of Ig light chain gene rearrangement in heavy chain-deficient B cells by activated Ras. Proc. Natl. Acad. Sci. USA 96:2239–43 Kee BL, Rivera RR, Murre C. 2001. Id3 inhibits B lymphocyte progenitor growth and survival in response to TGF-β. Nat. Immunol. 2:242–47 Schlissel M, Voronova A, Baltimore D. 1991. Helix-loop-helix transcription factor E47 activates germ-line immunoglobulin heavy-chain gene transcription and rearrangement in a pre-T-cell line. Genes Dev. 5:1367–76 Romanow WJ, Langerak AW, Goebel P, Wolvers-Tettero IL, van Dongen JJ, Feeney AJ, Murre C. 2000. E2A and EBF act in synergy with the V(D)J recombinase to generate a diverse immunoglobulin repertoire in nonlymphoid cells. Mol. Cell 5:343–53 Goebel P, Janney N, Valenzuela JP,

1 Feb 2002

10:17

320

66.

Annu. Rev. Immunol. 2002.20:301-322. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

67. 68.

69.

70.

71.

72.

73.

74.

75.

76.

AR

QUONG

AR152-11.tex

¥

AR152-11.SGM

ROMANOW

¥

LaTeX2e(2001/05/10)

P1: GJC

MURRE

Romanow WJ, Murre C, Feeney AJ. 2001. Localized gene-specific induction of accessibility to V(D)J recombination induced by E2A and early B cell factor in nonlymphoid cells. J. Exp. Med. 194:645– 56 Shimizu A, Honjo T. 1984. Immunoglobulin class switching. Cell 36:801–3 MacLennan IC. 1994. Germinal centers. Annu. Rev. Immunol. 12:117–39 Han S, Zheng B, Takahashi Y, Kelsoe G. 1997. Distinctive characteristics of germinal center B cells. Semin. Immunol. 9: 255–60 Wabl M, Cascalho M, Steinberg C. 1999. Hypermutation in antibody affinity maturation. Curr. Opin. Immunol. 11:186– 89 Quong MW, Harris DP, Swain SL, Murre C. 1999. E2A activity is induced during Bcell activation to promote immunoglobulin class switch recombination. EMBO J. 18:6307–18 Goldfarb AN, Flores JP, Lewandowska K. 1996. Involvement of the E2A basic helix-loop-helix protein in immunoglobulin heavy chain class switching. Mol. Immunol. 33:947–56 Pan L, Sato S, Frederick JP, Sun XH, Zhuang Y. 1999. Impaired immune responses and B-cell proliferation in mice lacking the Id3 gene. Mol. Cell. Biol. 19: 5969–80 Deed RW, Hara E, Atherton GT, Peters G, Norton JD. 1997. Regulation of Id3 cell cycle function by Cdk-2-dependent phosphorylation. Mol. Cell. Biol. 17: 6815–21 Z´un˜ iga-Pfl¨ucker JC, Lenardo MJ. 1996. Regulation of thymocyte development from immature progenitors. Curr. Opin. Immunol. 8:215–24 Fehling HJ, von Boehmer H. 1997. Early αβ T cell development in the thymus of normal and genetically altered mice. Curr. Opin. Immunol. 9:263–75 Marrack P, Kappler J. 1997. Positive selection of thymocytes bearing alpha beta

77.

78.

79.

80.

81.

82.

83.

84.

85.

T cell receptors. Curr. Opin. Immunol. 9:250–55 Killeen N, Irving BA, Pippig S, Zingler K. 1998. Signaling checkpoints during the development of T lymphocytes. Curr. Opin. Immunol. 10:360–67 Williams NS, Klem J, Puzanov IJ, Sivakumar PV, Schatzle JD, Bennett M, Kumar V. 1998. Natural killer cell differentiation: insights from knockout and transgenic mouse models and in vitro systems. Immunol. Rev. 165:47–61 Carlyle JR, Z´un˜ iga-Pfl¨ucker JC. 1998. Lineage commitment and differentiation of T and natural killer lymphocytes in the fetal mouse. Immunol. Rev. 165:63–74 Spits H, Blom B, Jaleco AC, Weijer K, Verschuren MC, van Dongen JJ, Heemskerk MH, Res PC. 1998. Early stages in the development of human T, natural killer and thymic dendritic cells. Immunol. Rev. 165:75–86 Ikawa T, Fujimoto S, Kawamoto H, Katsura Y, Yokota Y. 2001. Commitment to natural killer cells requires the helix-loophelix inhibitor Id2. Proc. Natl. Acad. Sci. USA 98:5164–69 Heemskerk MH, Blom B, Nolan G, Stegmann AP, Bakker AQ, Weijer K, Res PC, Spits H. 1997. Inhibition of T cell and promotion of natural killer cell development by the dominant negative helix loop helix factor Id3. J. Exp. Med. 186:1597–602 Blom B, Heemskerk MH, Verschuren MC, van Dongen JJ, Stegmann AP, Bakker AQ, Couwenberg F, Res PC, Spits H. 1999. Disruption of αβ but not of γ δ T cell development by overexpression of the helix-loop-helix protein Id3 in committed T cell progenitors. EMBO J. 18:2793–802 Engel I, Johns C, Bain G, Rivera RR, Murre C. 2001. Early thymocyte development is regulated by modulation of E2A protein activity. J. Exp. Med. 194:733– 45 Bain G, Romanow WJ, Albers K, Havran WL, Murre C. 1999. Positive and negative regulation of V(D)J recombination

1 Feb 2002

10:17

AR

AR152-11.tex

AR152-11.SGM

LaTeX2e(2001/05/10)

P1: GJC

E PROTEINS IN LYMPHOID DEVELOPMENT

86.

Annu. Rev. Immunol. 2002.20:301-322. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

87.

88.

89.

90.

91.

92.

93. 94.

95. 96.

by the E2A proteins. J. Exp. Med. 189: 289–300 Ghosh JK, Romanow WJ, Murre C. 2001. Induction of a diverse T cell receptor γ /δ repertoire by the helix-loop-helix proteins E2A and HEB in nonlymphoid cells. J. Exp. Med. 193:769–76 Langerak AW, Wolvers-Tettero ILM, van Gastel-Mol EJ, Oud MECM, van Dongen JJM. 2001. Basic helix-loop-helix proteins E2A and HEB induce immature Tcell receptor rearrangements in nonlymphoid cells. Blood 98:2456–65 Kim D, Peng XC, Sun XH. 1999. Massive apoptosis of thymocytes in T-celldeficient Id1 transgenic mice. Mol. Cell. Biol. 19:8240–53 Barndt RJ, Dai M, Zhuang Y. 2000. Functions of E2A-HEB heterodimers in T-cell development revealed by a dominant negative mutation of HEB. Mol. Cell. Biol. 20:6677–85 Kruisbeek AM, Haks MC, Carleton M, Michie AM, Z´un˜ iga-Pfl¨ucker JC, Wiest DL. 2000. Branching out to gain control: how the pre-TCR is linked to multiple functions. Immunol. Today 21:637– 44 Herblot S, Steff AM, Hugo P, Aplan PD, Hoang T. 2000. SCL and LMO1 alter thymocyte differentiation: inhibition of E2AHEB function and pre-T α chain expression. Nat. Immunol. 1:138–44 Barndt R, Dai MF, Zhuang Y. 1999. A novel role for HEB downstream or parallel to the pre-TCR signaling pathway during αβ thymopoiesis. J. Immunol. 163:3331– 43 von Boehmer H. 1994. Positive selection of lymphocytes. Cell 76:219–28 Robey E, Fowlkes BJ. 1994. Selective events in T cell development. Annu. Rev. Immunol. 12:675–705 Jameson SC, Bevan MJ. 1998. T-cell selection. Curr. Opin. Immunol. 10:214–19 Bain G, Quong MW, Soloff RS, Hedrick SM, Murre C. 1999. Thymocyte maturation is regulated by the activity of the

97.

98.

99.

100.

101.

102.

103.

104.

105.

106.

321

helix-loop-helix protein, E47. J. Exp. Med. 190:1605–16 Crompton T, Gilmour KC, Owen MJ. 1996. The MAP kinase pathway controls differentiation from double-negative to double-positive thymocyte. Cell 86:243– 51 Alberola-Ila J, Forbush KA, Seger R, Krebs EG, Perlmutter RM. 1995. Selective requirement for MAP kinase activation in thymocyte differentiation. Nature 373:620–23 Pages G, Guerin S, Grall D, Bonino FD, Smith A, Anjuere F, Auberger P, Pouyssegur J. 1999. Defective thymocyte maturation in p44 MAP kinase (Erk 1) knockout mice. Science 286:1374–77 Bain G, Cravatt CB, Loomans C, Alberola-Ila J, Hedrick SM, Murre C. 2001. Regulation of the helix-loop-helix proteins, E2A and Id3, by the Ras-ERK MAPK cascade. Nat. Immunol. 2:165–71 Sharp LL, Schwarz DA, Bott CM, Marshall CJ, Hedrick SM. 1997. The influence of the MAPK pathway on T cell lineage commitment. Immunity 7:609–18 Hollnagel A, Oehlmann V, Heymer J, Ruther U, Nordheim A. 1999. Id genes are direct targets of bone morphogenetic protein induction in embryonic stem cells. J. Biol. Chem. V274:19,838–45 Lasorella A, Noseda M, Beyna M, Iavarone A. 2000. Id2 is a retinoblastoma protein target and mediates signalling by Myc oncoproteins. Nature 407:592–98 Radtke F, Wilson A, Stark G, Bauer M, van Meerwijk J, MacDonald HR, Aguet M. 1999. Deficient T cell fate specification in mice with an induced inactivation of Notch1. Immunity 10:547–58 Pear WS, Aster JC, Scott ML, Hasserjian RP, Soffer B, Sklar J, Baltimore D. 1996. Exclusive development of T cell neoplasms in mice transplanted with bone marrow expressing activated Notch alleles. J. Exp. Med. 183:2283–91 Pui JC, Allman D, Xu LW, DeRocco S, Karnell FG, Bakkour S, Lee JY, Kadesch

1 Feb 2002

10:17

322

107.

Annu. Rev. Immunol. 2002.20:301-322. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

108.

109.

110.

111.

112.

113.

114.

115.

AR

QUONG

AR152-11.tex

¥

AR152-11.SGM

ROMANOW

¥

LaTeX2e(2001/05/10)

P1: GJC

MURRE

T, Hardy RR, Aster JC, Pear WS. 1999. Notch1 expression in early lymphopoiesis influences B versus T lineage determination. Immunity 11:299–308 Ordentlich P, Lin A, Shen CP, Blaumueller C, Matsuno K, Artavanis-Tsakonas S, Kadesch T. 1998. Notch inhibition of E47 supports the existence of a novel signaling pathway. Mol. Cell. Biol. 18:2230–39 Engel I, Murre C. 1999. Ectopic expression of E47 or E12 promotes the death of E2A-deficient lymphomas. Proc. Natl. Acad. Sci. USA 96:996–1001 Park ST, Nolan GP, Sun XH. 1999. Growth inhibition and apoptosis due to restoration of E2A activity in T cell acute lymphoblastic leukemia cells. J. Exp. Med. 189:501–8 Yan W, Young AZ, Soares VC, Kelley R, Benezra R, Zhuang Y. 1997. High incidence of T-cell tumors in E2A-null mice and E2A/Id1 double-knockout mice. Mol. Cell. Biol. 17:7317–27 Morrow MA, Mayer EW, Perez CA, Adlam M, Siu G. 1999. Overexpression of the Helix-Loop-Helix protein Id2 blocks T cell development at multiple stages. Mol. Immunol. 36:491–503 Kamps MP, Murre C, Sun XH, Baltimore D. 1990. A new homeobox gene contributes the DNA binding domain of the t(1;19) translocation protein in pre-B ALL. Cell 60:547–55 Inaba T, Roberts WM, Shapiro LH, Jolly KW, Raimondi SC, Smith SD, Look AT. 1992. Fusion of the leucine zipper gene HLF to the E2A gene in human acute Blineage leukemia. Science 257:531–34 Abraham KM, Levin SD, Marth JD, Forbush KA, Perlmutter RM. 1991. Thymic tumorigenesis induced by overexpression of p56lck. Proc. Natl. Acad. Sci. USA 88:3977–81 Makris A, Patriotis C, Bear SE, Tsichlis PN. 1993. Genomic organization and expression of Tpl-2 in normal cells and

116.

117.

118.

119.

120.

121.

122.

123.

Moloney murine leukemia virus-induced rat T-cell lymphomas: activation by provirus insertion. J. Virol. 67:4283–89 Patriotis C, Makris A, Bear SE, Tsichlis PN. 1993. Tumor progression locus 2 (Tpl-2) encodes a protein kinase involved in the progression of rodent T-cell lymphomas and in T-cell activation. Proc. Natl. Acad. Sci. USA 90:2251–55 Salmeron A, Ahmad TB, Carlile GW, Pappin D, Narsimhan RP, Ley SC. 1996. Activation of MEK-1 and SEK-1 by Tpl2 proto-oncoprotein, a novel MAP kinase kinase kinase. EMBO J. 15:817–26 Hagemann D, Troppmair J, Rapp U. 1999. Cot proto-oncoprotein activates the dual specificity kinases MEK-1 and SEK1 and induces differentiation of PC12 cells. Oncogene 18:1391–400 Kleeff J, Ishiwata T, Friess H, Buchler MW, Israel MA, Korc M. 1998. The helix-loop-helix protein Id2 is overexpressed in human pancreatic cancer. Cancer Res. 58:3769–72 Maruyama H, Kleeff J, Wildi S, Friess H, Buchler MW, Israel MA, Korc M. 1999. Id-1 and Id-2 are overexpressed in pancreatic cancer and in dysplastic lesions in chronic pancreatitis. Am. J. Pathol. 155:815–22 Weijzen S, Velders MP, Kast WM. 1999. Modulation of the immune response and tumor growth by activated Ras. Leukemia 13:502–13 Neufeld B, Grosse-Wilde A, Hoffmeyer A, Jordan BWM, Chen P, Dinev D, Ludwig S, Rapp UR. 2000. Serine/threonine kinases 3pK and MAPK-activated protein kinase 2 interact with the basic helix-loophelix transcription factor E47 and repress its transcriptional activity. J. Biol. Chem. 275:20,239–42 Chu C, Kohtz DS. 2001. Identification of the E2A gene products as regulatory targets of the G1 cyclin-dependent kinases. J. Biol. Chem. 276:8524–34

P1: FRK

February 1, 2002

12:47

Annual Reviews

AR152-FM

Annual Review of Immunology Volume 20, 2002

Annu. Rev. Immunol. 2002.20:301-322. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

CONTENTS FRONTISPIECE, Charles A. Janeway, Jr. A TRIP THROUGH MY LIFE WITH AN IMMUNOLOGICAL THEME, Charles A. Janeway, Jr.

xiv 1

THE B7 FAMILY OF LIGANDS AND ITS RECEPTORS: NEW PATHWAYS FOR COSTIMULATION AND INHIBITION OF IMMUNE RESPONSES, Beatriz M. Carreno and Mary Collins

29

MAP KINASES IN THE IMMUNE RESPONSE, Chen Dong, Roger J. Davis, and Richard A. Flavell

PROSPECTS FOR VACCINE PROTECTION AGAINST HIV-1 INFECTION AND AIDS, Norman L. Letvin, Dan H. Barouch, and David C. Montefiori T CELL RESPONSE IN EXPERIMENTAL AUTOIMMUNE ENCEPHALOMYELITIS (EAE): ROLE OF SELF AND CROSS-REACTIVE ANTIGENS IN SHAPING, TUNING, AND REGULATING THE AUTOPATHOGENIC T CELL REPERTOIRE, Vijay K. Kuchroo, Ana C. Anderson, Hanspeter Waldner, Markus Munder, Estelle Bettelli, and Lindsay B. Nicholson

55 73

101

NEUROENDOCRINE REGULATION OF IMMUNITY, Jeanette I. Webster, Leonardo Tonelli, and Esther M. Sternberg

125

MOLECULAR MECHANISM OF CLASS SWITCH RECOMBINATION: LINKAGE WITH SOMATIC HYPERMUTATION, Tasuku Honjo, Kazuo Kinoshita, and Masamichi Muramatsu

165

INNATE IMMUNE RECOGNITION, Charles A. Janeway, Jr. and Ruslan Medzhitov

KIR: DIVERSE, RAPIDLY EVOLVING RECEPTORS OF INNATE AND ADAPTIVE IMMUNITY, Carlos Vilches and Peter Parham ORIGINS AND FUNCTIONS OF B-1 CELLS WITH NOTES ON THE ROLE OF CD5, Robert Berland and Henry H. Wortis E PROTEIN FUNCTION IN LYMPHOCYTE DEVELOPMENT, Melanie W. Quong, William J. Romanow, and Cornelis Murre

197 217 253 301

LYMPHOCYTE-MEDIATED CYTOTOXICITY, John H. Russell and Timothy J. Ley

SIGNAL TRANSDUCTION MEDIATED BY THE T CELL ANTIGEN x

323

P1: FRK

Annu. Rev. Immunol. 2002.20:301-322. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

February 1, 2002

12:47

Annual Reviews

AR152-FM

CONTENTS

xi

RECEPTOR: THE ROLE OF ADAPTER PROTEINS, Lawrence E. Samelson INTERACTION OF HEAT SHOCK PROTEINS WITH PEPTIDES AND ANTIGEN PRESENTING CELLS: CHAPERONING OF THE INNATE AND ADAPTIVE IMMUNE RESPONSES, Pramod Srivastava CHROMATIN STRUCTURE AND GENE REGULATION IN THE IMMUNE SYSTEM, Stephen T. Smale and Amanda G. Fisher PRODUCING NATURE’S GENE-CHIPS: THE GENERATION OF PEPTIDES FOR DISPLAY BY MHC CLASS I MOLECULES, Nilabh Shastri, Susan

371

Schwab, and Thomas Serwold

395 427

463

THE IMMUNOLOGY OF MUCOSAL MODELS OF INFLAMMATION, Warren Strober, Ivan J. Fuss, and Richard S. Blumberg T CELL MEMORY, Jonathan Sprent and Charles D. Surh

GENETIC DISSECTION OF IMMUNITY TO MYCOBACTERIA: THE HUMAN MODEL, Jean-Laurent Casanova and Laurent Abel ANTIGEN PRESENTATION AND T CELL STIMULATION BY DENDRITIC CELLS, Pierre Guermonprez, Jenny Valladeau, Laurence Zitvogel, Clotilde Th´ery, and Sebastian Amigorena

495 551 581

621

NEGATIVE REGULATION OF IMMUNORECEPTOR SIGNALING, Andr´e Veillette, Sylvain Latour, and Dominique Davidson

669

CPG MOTIFS IN BACTERIAL DNA AND THEIR IMMUNE EFFECTS, Arthur M. Krieg

PROTEIN KINASE Cθ

709 IN

T CELL ACTIVATION, Noah Isakov and Amnon

Altman

RANK-L AND RANK: T CELLS, BONE LOSS, AND MAMMALIAN EVOLUTION, Lars E. Theill, William J. Boyle, and Josef M. Penninger PHAGOCYTOSIS OF MICROBES: COMPLEXITY IN ACTION, David M. Underhill and Adrian Ozinsky

761 795 825

STRUCTURE AND FUNCTION OF NATURAL KILLER CELL RECEPTORS: MULTIPLE MOLECULAR SOLUTIONS TO SELF, NONSELF DISCRIMINATION, Kannan Natarajan, Nazzareno Dimasi, Jian Wang, Roy A. Mariuzza, and David H. Margulies

853

INDEXES Subject Index Cumulative Index of Contributing Authors, Volumes 1–20 Cumulative Index of Chapter Titles, Volumes 1–20

ERRATA An online log of corrections to Annual Review of Immunology chapters (if any, 1997 to the present) may be found at http://immunol.annualreviews.org/

887 915 925

8 Feb 2002

9:54

AR

AR152-12.tex

AR152-12.SGM

LaTeX2e(2001/05/10) P1: GJC 10.1146/annurev.immunol.20.100201.131730

Annu. Rev. Immunol. 2002. 20:323–70 DOI: 10.1146/annurev.immunol.20.100201.131730 c 2002 by Annual Reviews. All rights reserved Copyright °

LYMPHOCYTE-MEDIATED CYTOTOXICITY John H. Russell1 and Timothy J. Ley2 Department of Molecular Biology and Pharmacology, 2Departments of Medicine and Genetics, Siteman Cancer Center, Washington University School of Medicine, St. Louis, Missouri 63110; e-mail: [email protected], [email protected]

Annu. Rev. Immunol. 2002.20:323-370. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

1

Key Words granule exocytosis, Fas, apoptosis, perforin, granzymes ■ Abstract Virtually all of the measurable cell-mediated cytotoxicity delivered by cytotoxic T lymphocytes and natural killer cells comes from either the granule exocytosis pathway or the Fas pathway. The granule exocytosis pathway utilizes perforin to traffic the granzymes to appropriate locations in target cells, where they cleave critical substrates that initiate DNA fragmentation and apoptosis; granzymes A and B induce death via alternate, nonoverlapping pathways. The Fas/FasL system is responsible for activation-induced cell death but also plays an important role in lymphocyte-mediated killing under certain circumstances. The interplay between these two cytotoxic systems provides opportunities for therapeutic interventions to control autoimmune diseases and graft vs. host disease, but oversuppression of these pathways may also lead to increased viral susceptibility and/or decreased tumor cell killing.

INTRODUCTION Mechanisms of lymphocyte-mediated cytotoxicity were last reviewed in this series by Kagi et al. (1). Much of our understanding of cytotoxic pathways has come from experiments in cultured cells. At the time of the previous review, it had been established that the two dominant mechanisms of contact-dependent, lymphocytemediated cytotoxicity were the granule exocytosis and the Fas pathways (2), and animals deficient in each of these pathways had just recently been identified or produced (3–5). Here we focus on two elements: the biological significance of these pathways in various in vivo settings, and the molecular components of these pathways. A better understanding of these two elements is necessary to provide targets for rational therapeutic intervention in pathological conditions.

LYMPHOCYTE-MEDIATED KILLING AS AN INTERNAL EVENT IN THE TARGET CELL As early as the 1950s, morphological distinctions had been made between targets attacked by lymphocytes vs. targets attacked by antibody and complement, especially when one compared effects on nuclear structure (6). These differences 0732-0582/02/0407-0323$14.00

323

5 Feb 2002

12:52

Annu. Rev. Immunol. 2002.20:323-370. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

324

AR

AR152-12.tex

RUSSELL

¥

AR152-12.SGM

LaTeX2e(2001/05/10)

P1: GJC

LEY

were subsequently confirmed using isotope release assays that distinguished between events affecting the plasma membrane and those affecting the nucleus. These experiments demonstrated that antibody and complement damage was restricted to the plasma membrane, while lymphocytes produced a more general internal disintegration of the target cell, including the nucleus (7, 8). Such experiments suggested that target cells may play an active role in their own destruction, a notion strengthened by the observation that the severity of the nuclear lesion inflicted by the same cytotoxic lymphocyte (CTL) varied with the lineage of the target cell (9). It is likely that these differences reflect differences in the levels of effector caspases and caspase and/or of granzyme substrates in cells of different lineages.

TWO PATHWAYS OF LYMPHOCYTE-MEDIATED KILLING In the earlier review, Kagi et al. suggested that lymphocyte-mediated killing could be confined to two pathways, the perforin/granzyme-mediated and the Fasmediated pathways (1). Emerging evidence suggests that, although this interpretation is true to a first approximation, it may be more useful to consider the two mechanisms as those initiated by FADD through a target cell receptor and those that require perforin. Based on genetically deficient animals, Fas is the most physiologically important receptor initiating death through the recruitment of FADD and caspase 8, but other members of the TNF receptor pathway, including TNFR1 and TRAILR, converge on FADD as well (10–13). The emerging evidence discussed below suggests that the perforin/granzyme pathway is not a single pathway, but rather a series of parallel pathways that depend on the particular granzyme or spectrum of granzymes expressed in a given effector lineage or activation state. Several important similarities and differences appear between the two pathways. The perforin-dependent pathway is dominant in CD8+ CTL and natural killer (NK) cells (2). In NK cells, granules are preformed, although NK activity can be increased by cytokines like IL-2 and IFN-γ . NK cells are thus constitutively armed and can kill within minutes of the first stimulation of activating receptors, but they generally do not proliferate significantly in response to this stimulation. Thus, NK cells are part of the innate immune system and have a rapid response to challenge, but limited capacity for antigen/viral loads. In contrast, naive CD8+ CTL precursors have no cytotoxic activity and must undergo an activation process requiring 1–3 days for maximal activity. This activation process requires TCR-stimulated induction of cytokine receptors (e.g., IL-2 and IL-6), which then induce the expression of granule components, including perforin and granzymes. The same signals that activate CD8+ cells also stimulate their proliferation. As recently reviewed (14), virus-specific CD8+ can expand several orders of magnitude in response to viral infection. Thus, CD8+ cells are

5 Feb 2002

12:52

AR

AR152-12.tex

AR152-12.SGM

LaTeX2e(2001/05/10)

P1: GJC

Annu. Rev. Immunol. 2002.20:323-370. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

LYMPHOCYTE-MEDIATED CYTOTOXICITY

325

part of the adaptive immune system with a slow response, but a high capacity for viral load. The activated CD8+ cells are then armed for their next encounter with antigenbearing target cells. When granules are present, the killer cell reorients its granules to the region of receptor activation (NK or TCR) and releases the granule components into the region of contact between the killer and the target. This region of contact is initiated by the receptor but maintained by adhesion molecules (e.g., LFA-1 and ICAM-1). CD8+ cells can kill multiple cells by reorienting their granules to another region of contact, but there is evidence that NK cells must rearm themselves in response to IL-2 before they are effective against new targets (15). The vectorial release of granules to the region of contact between the killer and target cells provides one mechanism of exquisite specificity of the perforin-initiated pathway (16). A second mechanism is that perforin reacts nonspecifically with lipids in the presence of Ca++ (17) and therefore is rapidly inactivated by Ca++ and lipids in the extracellular space. The mechanism of killing by perforin and the granzymes is discussed in more detail below. The FADD pathway (see below) is similar in that maximal expression of the TNF family of ligands requires an initial activation of the effector cell (18). This pathway appears to be active in all killer cell lineages but most important for CD4+ cells, especially those of the Th1 phenotype (19). An important difference between the FADD and perforin-initiated pathways is the speed of the cytotoxic event. Once formed, granules can be reoriented and released within minutes of TCR stimulation (16). In contrast, very little ligand (e.g., FasL) is stored, even in activated cells. Therefore, maximal activity requires the induction of new ligand over a 1–2 h period after TCR stimulation. Induction of new ligand continues as long as there is TCR stimulation and FasL is cleared from the surface either by proteolysis or endocytosis, depending on the effector cell. This clearance of ligand from the surface occurs with a half-life of 2–3 h (20). The long half-life of ligand on the surface allows effector cells to continue to display cytotoxic activity in the absence of TCR stimulation and leads to the phenomenon of bystander killing, which means that cells in the area that express the appropriate receptor (e.g., Fas) can be killed even though they do not express the antigen recognized by the TCR (21). This can be especially important for CD4+ cells whose class II restricting element is not widely expressed on a variety of cellular lineages. Thus, the FADD pathway can be much more promiscuous than the perforin-initiated pathway. It is clear that cultured human and mouse CD4+ and CD8+ cell lines can use both the Fas and perforin/granzyme pathways (19, 22, 23). However, in vivo transplantation and graft vs. host disease (GvHD) experiments across MHC class I or II (CD8+ and CD4+ effectors, respectively) suggest that the perforin/granzyme pathway dominates the class I elimination pathway and that Fas/FasL dominates class II elimination (24, 25). Perforin deficiency has the major phenotype in an in vivo model of NK-mediated tumor rejection (26).

5 Feb 2002

12:52

326

AR

AR152-12.tex

RUSSELL

¥

AR152-12.SGM

LaTeX2e(2001/05/10)

P1: GJC

LEY

Annu. Rev. Immunol. 2002.20:323-370. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

APOPTOTIC PATHWAYS The last 15 years of research in numerous laboratories have demonstrated that vertebrate cell populations have two ways of maintaining homeostasis (see Figure 1). The first is a pathway that is evolutionarily conserved in multicellular organisms as distinct as C. elegans, D. melanogaster, and H. sapiens. This pathway is known as the intrinsic pathway and is important for the appropriate development of a variety of organ systems (especially the nervous system) and the elimination of cells that have developed abnormally or with genetic errors. The other pathway, known as the extrinsic pathway (the FADD pathway discussed above), may be unique to vertebrates. Both of these pathways have recently been reviewed (27, 28). Both pathways are based on protein-protein interaction domains that lead to the activation of a cascade of proteases with the unusual properties of a cysteine residue in the active site, and a specificity for cleavage at aspartic acid residues (caspases). The activation of these caspases is modulated physiologically by functional dominant

Figure 1 Extrinsic and intrinsic signaling pathways of apoptosis. Two pathways exist for the correct development and homeostasis of organs and cell populations. The extrinsic pathway begins with ligation of a membrane death receptor (e.g., Fas) and is transduced through a series of protein-protein interaction domains culminating in the activation of unique proteases (caspases). The intrinsic pathway is triggered by a lack of trophic receptor stimulation, DNA damage, glucocorticoids, or inappropriate loss of contact with neighboring cells. This ultimately leads to a loss of mitochondrial function and cytochrome c release, which in turn activates the Apaf-1/caspase 9 complex (“apoptosome”), causing further caspase activation.

5 Feb 2002

12:52

AR

AR152-12.tex

AR152-12.SGM

LaTeX2e(2001/05/10)

P1: GJC

LYMPHOCYTE-MEDIATED CYTOTOXICITY

327

negative caspases (e.g., FLICE-like inhibitory protein, c-FLIP/Casper) or caspase binding proteins (inhibitors of apoptosis, IAPs). The substrates of the various caspases (29) are responsible for the morphological (cytoskeletal, nuclear membrane breakdown) and biochemical (DNA laddering) changes associated with apoptosis. As is discussed in more detail below, the best understood granzyme (B) can act on both caspases and caspase substrates in both the intrinsic and extrinsic pathway.

Annu. Rev. Immunol. 2002.20:323-370. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

PROTECTION FROM INTRACELLULAR PARASITES CTL of the MHC class I–restricted, CD8+ lineage are uniquely selected to control intracellular pathogens because of the loading of the MHC class I molecule with peptides derived from intracellular proteins. Early experiments by Zinkernagel demonstrated that CTL could kill viral-infected cells before new viral antigens could be detected on the surface by antibodies (30). The availability of null or functional null mutations in the Fas and perforin/granzyme pathways has allowed a first approximation analysis of the importance of these pathways in the control of viral infections (see Table 1). The results of these experiments indicate that the perforin/granzyme pathway is absolutely essential for the control of some viral infections, has a partial role in some, and appears to play no role in others. In contrast, the Fas pathway plays little role in the clearance of viruses unless the perforin pathway is also inactivated (31, 32); Fas may also play a role in clearing some reservoirs of persistent viral infection (see below) (33). The situation with a persistent, neurotropic variant of murine hepatitis virus is particularly interesting. Perforin deficiency significantly delays viral clearance from the CNS (34). Fas deficiency had no effect on the role of viral clearance, but elimination of both Fas and perforin led to uncontrolled infection (32). However, the virus also could not be cleared from mice deficient in IFN-γ . Rather, these mice developed increased CD8+ infiltrates, and virus was cleared from microglia and astrocytes, but not from oligodendrocytes (33). The role of IFN-γ in selectively limiting viral replication in oligodendrocytes has not been established, but one might speculate that IFN-γ could affect the sensitivity of the oligodendrocytes to CTL by increasing either viral peptide MHC class I or Fas expression, rendering this normally resistant cell sensitive to the action of CTL. The initial experiments with perforin-deficient mice suggested that perforin was important in controlling noncytopathic viral infections, but not cytopathic ones (35). However, this is clearly an oversimplification because the perforin/granzyme pathway is involved in protection from some cytopathic viruses (36). With other viruses, the type of viral pathology may depend upon the cell type infected (34). Most of the viral infection experiments have been performed only with perforindeficient mice, so it is unclear whether different granzyme pathways may be more important in some viral infections than others. Similarly, only a few of the experiments have attempted to identify the relative importance of different perforin-expressing effectors essential for controlling the virus (e.g., CTL vs. NK). Analysis of the role of specific granzymes in viral protection may point to

5 Feb 2002

12:52

328

AR

AR152-12.tex

RUSSELL

¥

AR152-12.SGM

LaTeX2e(2001/05/10)

P1: GJC

LEY

Annu. Rev. Immunol. 2002.20:323-370. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

TABLE 1 Analysis of apoptotic pathways in protection from viral infection using null mutations Virus

Perforin

Fas

Notes (with references)

Lymphocytic choriomeningitis virus

+

−

Complete protection by perforin deficiency (35)

Vesicular stomatitis virus

−

−

(35)

Semliki forest virus

−

−

(35)

Vaccinia

−

−

(35)

Ectromelia

+

NT

(36)

Cowpox

−

NT

(36)

Theiler’s virus

+

−

Acute infection leads to death (261) Also MHC class I-dependent and decreased neurologic symptoms (262)

MAIDS

+

NT

Also MHC class I dependent (263)

−

No effect of perforin deficiency on viral clearance from cornea, but decreased keratitis in the absence of perforin (38) or perforin protection from lethal infection, but no difference in viral clearance from the eye (37)

+

Herpes simplex-1

−/

Coxsackie B3

−

NT

No role of perforin for viral clearance, but decreased myocarditis in the absence of perforin (79)

Neurotropic hepatitis virus

Partial

Only in the absence of perforin

Absence of perforin delays viral clearance (34) IFN-γ required for complete clearance (33)

Influenza

Partial

“

Role for perforin or Fas is only evident when CD4 helper cells are depleted (31)

Adenovirus vectors

−

+

Primarily dependent on TNFR1, which is required for monocytic infiltrate (39)

Murine cytomegalovirus

+

−

Perforin (264) but not Fas is required for viral clearance, but Fas is important in limiting chronic inflammatory disease (265)

γ Herpes

−

NT

(266)

/−

specific effector populations. Using mice deficient for specific granzymes to analyze protection against specific viruses will also lead to a better understanding of the interactions between viruses and the immune system that lead to the selection of viral inhibitors of apoptosis, and the parallel pathways in the immune system that exist to bypass those inhibitors.

5 Feb 2002

12:52

AR

AR152-12.tex

AR152-12.SGM

LaTeX2e(2001/05/10)

P1: GJC

Annu. Rev. Immunol. 2002.20:323-370. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

LYMPHOCYTE-MEDIATED CYTOTOXICITY

329

There are several difficulties with the viral experiments performed to date. The principal difficulty is that different viruses have specific viral-resistance genes scattered throughout the genome. In addition, many of the mutant mice are on mixed backgrounds between B6 and different 129 substrains; the viral resistance genetics of the 129 strains are poorly defined. Therefore, it is critically important to use large numbers of deficient animals (with littermates as controls) to eliminate possible effects of unknown background resistance genes. Many of the deficiency mutations are becoming available on the B6/B10 background, which minimizes the effects of resistance genes but limits the number of viruses that can be analyzed. Even when similar viruses and animal strains are used, results can vary, perhaps because of differences in viral dose or strain-specific pathogenicity (37, 38). Another difficulty, especially with deficiencies of the TNF receptor family, is that the mutations are pleiotropic and have profound effects on lymphocyte homeostasis and trafficking, as well as effector function. A good example is the finding that the TNFR1 mutation has the most dramatic effect on the clearance of adenoviral vectors from the liver through its effects on monocytic infiltration (39). TNFR1 has previously been demonstrated to play an important role in this process, through its effects on the induction of addressins and adhesion molecules (40). Intracellular bacteria have not been studied as extensively as viruses. Mice deficient in perforin (41) have a modestly attenuated primary response to Listeria infection and a somewhat more dramatic deficit in the secondary response. Mice deficient in Fas have a small defect, but the combination of Fas and perforin deficiency greatly prolongs the clearance time (42). However, all these deficient mice recover from severe infection. It appears that TNF (43) from a non-CD8 source (44) is crucial in limiting Listeriosis. Recent experiments suggest that CTL and NK granules from some species, including humans, contain proteins that can be used as defense against both intracellular and extracellular bacteria. These proteins have structural similarity to antimicrobial virulence factors of amoebae (45). The best characterized of these molecules is granulysin (46). This protein, originally identified as a T cell activation product, has the ability to kill tumor cells by caspase-dependent and caspase-independent mechanisms (47). However, it may be more important for its broad spectrum antimicrobial activity (48–50). Murine orthologues have not been found, so it has not been possible to test the physiology of granulysin through loss-of-function mutations.

CELL-MEDIATED CYTOTOXICITY IN IMMUNE REGULATION AND AUTOIMMUNE DISEASE Immune Regulation The role of Fas/FasL in lymphocyte homeostasis was clearly established with the recognition that functional null mutations in these proteins (lpr and gld, respectively) were associated with exacerbated autoimmune disease (4, 51). Evidence

5 Feb 2002

12:52

Annu. Rev. Immunol. 2002.20:323-370. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

330

AR

AR152-12.tex

RUSSELL

¥

AR152-12.SGM

LaTeX2e(2001/05/10)

P1: GJC

LEY

that cell-mediated cytotoxicity played a role in lymphocyte homeostasis was provided by Goodnow and colleagues in adoptive transfer experiments utilizing a transgenic model of B cell tolerance (52). These experiments implicated Fasdependent killing of anergized B cells by Th1 T cells. It was demonstrated that B cell sensitivity to Fas-dependent killing was regulated by signals from the BCR and CD40, another member of the TNF receptor family (28, 53). The Fas pathway is also important in T cell homeostasis, especially in chronic exposure to self-antigens, while the intrinsic pathway is more important after acute immunization (54). It is unclear whether Fas/FasL-dependent elimination of chronically activated cells is accomplished by one T cell killing another (fratricide) or through a specialized suicide process. There is evidence for both mechanisms in culture, but whether only one or both occur in vivo is difficult to determine. The Fas system plays little, if any, role in negative selection in the thymus (55–57). A number of reports regarding a potential role of the granule exocytosis pathway for lymphocyte homeostasis were recently reviewed by de Saint Basile and Fischer, who suggest that defects in this pathway are primarily responsible for the hemophagocytic syndromes (58). Clearly, some mutations that affect granule formation and trafficking in many cell lineages have similar phenotypes in both mice and humans (59–61). However, animals deficient for perforin have lymphocyte populations that are phenotypically normal. Perforin-deficient animals do develop a syndrome with hematophagocytosis when they are infected with viruses that they cannot clear (62, 63), but it is unclear whether the increase in the number of activated CD8+ cells and macrophages is a failure of lymphocyte homeostasis or the consequence of chronic inflammation due to persistent viral infection. A secondary role for the granule exocytosis pathway in lymphocyte hemostasis is unmasked by an exacerbation of the lymphoproliferative phenotype when a null mutation in perforin is added to the lpr or gld mutations (64, 65). Similarly, deficiency in TNFR1 dramatically exacerbates both the lymphoproliferative disease and autoantibody production when combined with the lpr mutation (66). A selective role of an interaction between TNFR1, TNFR2, and Fas in mature CD8+ cell homeostasis has also been suggested (67–69). This hypothesis has produced modest results when tested in vivo (70, 71). Thus, the Fas system has a major role in lymphocyte homeostasis, along with other cytotoxic ligands, working through TNF receptors. The granule exocytosis pathway can also contribute. How much of this activity is mediated by classical killing of one cell by another, vs. cytokineassisted suicide, remains to be determined.

Autoimmune Disease The potential roles of the granule exocytosis and Fas pathways in autoimmune disease have been most extensively studied in induced or spontaneous diabetes and induced demyelinating disease (EAE). The clearest demonstration of a role for the perforin/granzyme pathway is in induced diabetes. In these models, viral proteins expressed as transgenes in the β cells of the islets target the β cells

5 Feb 2002

12:52

AR

AR152-12.tex

AR152-12.SGM

LaTeX2e(2001/05/10)

P1: GJC

Annu. Rev. Immunol. 2002.20:323-370. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

LYMPHOCYTE-MEDIATED CYTOTOXICITY

331

for attack upon viral infection. In both models where the disease is dependent upon CD8+ effectors, perforin deficiency eliminates or dramatically reduces the severity of the induced disease without affecting lymphocytic infiltration of the islets (72, 73). In the NOD model of spontaneous diabetes, there is evidence that both the perforin/granzyme and Fas pathways (as well as others) are important in pathogenesis (74–78). An interesting role for the perforin/granzyme pathway in virally induced autoimmune disease comes from studies on coxsackie virus-induced myocarditis (79) and HSV-1-induced keratitis (38). In both studies, perforin deficiency had no effect on viral clearance, but dramatically ameliorated the autoimmune sequelae. In the latter study, the role of lymphocytes for initiating the disease was confirmed by adoptive transfer. However, in both instances, it appears that the actual effectors in the pathogenesis are not the CTL, but other cells in the inflammatory infiltrate. It has been known for some time that the process of apoptosis, including that induced by CTL, can be important in the release of proinflammatory mediators (80). In these two viral models, the apoptosis caused by the CTL in response to the viral infection is not important in controlling the virus but may be important for recruiting a pathogenic inflammatory infiltrate. Perforin deficiency did not ameliorate the development of induced demyelinating disease (EAE) in the model where it was tested. If anything, it appeared to slightly exacerbate disease, perhaps reflecting its role in lymphocyte homeostasis (81). In contrast, Fas/FasL deficiency dramatically limits disease incidence and severity in some models (82–84), but it has little or no effect in others (81, 85, 86). Some of these differences in the dominant effector mechanism appear to be caused by genetic influences that are distinct from MHC regulation of disease susceptibility (87). Because the Fas system is also important for lymphocyte homeostasis, null mutations in Fas or FasL can also exacerbate demyelinating disease in models where Fas is not a dominant mechanism of pathogenesis (86, 88). The evidence that cytotoxic activity of lymphocytes is involved in each of the above instances comes from adoptive transfer experiments using lymphocytes that cannot initiate a specific apoptotic pathway (68, 73, 77, 84, 89). These models indicate that CTL can use their apoptotic mechanisms to initiate autoimmune disease. The potential role of perforin in human disease is difficult to address. Perforinand granzyme-containing cells have been identified in patients at sites of autoimmune inflammation (90, 91), but their role in pathogenesis has not been established. There is similar circumstantial evidence for an involvement of Fas in several human autoimmune pathologies (92, 93), including multiple sclerosis (94, 95). Whether FasL from lymphocytes is always the initiating factor is somewhat controversial, in part because many of the FasL reagents used for immunohistochemistry have not been specific (96). The experiments in mice provide proof of the concept that the FADD pathway and the perforin/granzyme pathways are effectors not only in protection from pathogens, but also in the pathogenesis of autoimmune disease. As discussed below, the biochemical pathways utilized by the different granzymes are only

5 Feb 2002

12:52

332

AR

AR152-12.tex

RUSSELL

¥

AR152-12.SGM

LaTeX2e(2001/05/10)

P1: GJC

LEY

now being elucidated. A better understanding of these pathways, coupled with the continued analysis of the FADD pathway, may provide unique targets for therapeutic intervention in specific pathological situations.

Annu. Rev. Immunol. 2002.20:323-370. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

AN OVERVIEW OF THE GRANULE EXOCYTOSIS PATHWAY AND HOW IT CAUSES CELL DEATH The granule exocytosis pathway was first proposed in this journal in 1985 (97). When T cells receive specific signals to activate and proliferate via the T cell receptor, transcriptional mechanisms are activated that lead to the production of cytotoxic granules and their constituent proteins (98). In the case of NK cells, these granules are preformed during NK cell development, perhaps at an NK precursor stage that is not morphologically identifiable. Within a day after T cell activation, granules begin to be synthesized, along with perforin and granzymes and other granule components. These granules then reside in the cytoplasm of the cell, where they await further instructions. Upon target cell identification and conjugation, specific signals are generated in the effector lymphocyte that cause the granules to migrate by vector to the site of contact. At the cell surface, the granule fuses with the plasma cell membrane, and its contents are secreted into the tight intracellular junction formed between the two cells. There, in the presence of calcium, perforin polymerizes and enters the target cell membrane. In 1996, when this review was written (1), perforin was thought to form a channel through which the other granule proteins pass into the target cell cytoplasm. Perforin was therefore thought to act as an important gateway for granule proteins into target cells, and it was also thought that perforin could induce membrane damage that could ultimately lead to the death of the target cell. However, several early experiments suggested that perforin itself was not capable of causing target cell apoptosis unless molecules like the granzymes were also added (99–105). By 1996, it was clear that granzyme B was required for the rapid induction of target cell apoptosis by cytotoxic T cells and NK cells (106, 107). The roles of granzyme A and the many other granzymes found in cytotoxic granules were completely unknown. Granzyme B was known to prefer to cleave substrate proteins after an aspartic acid residue in the P1 position (108), and its similarity to the caspases was therefore well recognized. In fact, it was known that granzyme B could cleave procaspase 3 in vitro (109, 110) and that it was required for the cleavage of caspase 3 in target cells after CTLs delivered their granule load (111). It was widely believed that granzyme B killed by activating caspases in target cells, which then led to the cleavage of additional apoptotic substrates that led to DNA fragmentation and apoptosis. Because cytotoxic granules are self-contained death machines, these organelles have continued to be intensively studied over the past five years. A great deal of new information has come to light regarding every step in the pathway described above. Many of the notions described in the central dogma of 1996 have been

5 Feb 2002

12:52

AR

AR152-12.tex

AR152-12.SGM

LaTeX2e(2001/05/10)

P1: GJC

Annu. Rev. Immunol. 2002.20:323-370. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

LYMPHOCYTE-MEDIATED CYTOTOXICITY

333

shown to be flawed or simply incorrect. The roles of additional granule proteins in regulating the death process are beginning to be understood. Regardless, one thing has not changed. It is clear that activated CD8+ T cells and NK cells absolutely rely on the granule exocytosis pathway for their ability to induce target cell death and that this pathway is crucially important for the ability of CTL to induce target cell death in many biologically important scenarios, including allogeneic rejection, tumor cell killing, and the clearance of some viruses. In addition, a potential role for granule-mediated cytotoxicity in lymphocyte homeostasis has also evolved, even though this role was not recognized in 1996 (58, 112). The recent developments in our understanding of this complex pathway are outlined in the sections below.

GRANULE FORMATION AND CONSTITUENT GRANULE PROTEINS As noted above, the cytotoxic granules of activated cytotoxic T cells are synthesized de novo when the cell receives signals to proliferate and activate. Over the course of one to two days, the genes for perforin, granzyme A, granzyme B, and other granzymes are transcriptionally activated, and these newly synthesized proteins are then appropriately trafficked and assembled into the functional granule. After synthesis of the granzymes, important posttranslational modifications occur. First, the granzymes must be processed to assume an active conformation. The signal peptide is first removed by a signal peptidase, and then a short prosequence is removed by a second enzyme. In the case of granzymes A and B, the second enzyme is clearly dipeptidyl peptidase I (DPPI) (113–116). Mice deficient for DPPI synthesize normal amounts of granzymes A and B, but the enzymes are completely inactive since they are not processed normally at the N terminus (117). DPPI-deficient CTL are identical to granzyme A × B–deficient mice in their inability to induce apoptosis in allogeneic target cells. The precise compartment in which DPPI processes the granzymes into their mature forms is not known, but the granzymes are thought to exist in the granules as the active processed enzymes. The granzymes are also glycosylated and sorted by the Man-6-P receptor in the Golgi apparatus on their way to the specialized granules (118, 119). This modification of the granzymes may be important for their entry into target cells (see the section on the trafficking of perforin and granzymes below). Perforin is also thought to be modified posttranslationally to render it active. Uellner et al. (120) demonstrated that perforin is synthesized as a 70-kDa precursor that is cleaved at the carboxyl terminus to yield the active 60-kDa form. The enzyme that mediates this cleavage event has not yet been identified, but it may require an acidic environment (121). Accordingly, chloroquine, a lysosomatropic agent that raises the pH of the acidic compartment in which perforin is processed, may inhibit its maturation. Taylor et al. (122) have shown that chloroquine does indeed inhibit perforin activity in LAK cells, and that chloroquine-treated mice have defects in

5 Feb 2002

12:52

Annu. Rev. Immunol. 2002.20:323-370. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

334

AR

AR152-12.tex

RUSSELL

¥

AR152-12.SGM

LaTeX2e(2001/05/10)

P1: GJC

LEY

their ability to clear YAC-1 tumor cells from the lungs or to reject incompatible bone marrow transplants via traditional NK pathways. It is not yet known whether chloroquine acts solely to inhibit NK cell function by inhibiting perforin activation, or whether other effects may also contribute to the actions of the drug. Glycosaminoglycan complexes are also found in cytotoxic granules, including one proteoglycan called serglycin, which contains chondroitin 4-sulfate glycosaminoglycans. This molecule is thought to act as a scaffold for packaging the highly positively charged granzymes, and it may also act as a chaperone for the secreted proteases. Galvin et al. (123) have shown that granzyme B complexed with serglycin is proteolytically active and able to induce target cell apoptosis; they have also shown that granzyme B is actually secreted in a high-molecular-weight complex, presumably containing serglycin. The requirement of serglycin for the induction of target cell death is not yet understood, because recombinant granzyme B is fully capable of causing target cell death in reconstituted systems (124). Direct comparisons of the cytotoxic potential of serglycin-complexed granzymes and their recombinant counterparts have not yet been described. Calreticulin is also stored in cytotoxic granules. This molecule is a chaperone protein of the endoplasmic reticulum (ER) and is the only resident ER protein known to exist in CTL granules. Fraser et al. (125) have recently shown that calreticulin inhibits perforin-mediated target cell damage, but it does not do so by directly interacting with perforin, by sequestering calcium, or by inhibiting granzymes. The possibility therefore exists that calreticulin acts as a regulatory molecule that dampens the effect of perforin by stabilizing membranes to prevent excessive damage. Bossi & Griffiths (126) have shown that Fas ligand appears to be stored in the same cytotoxic granules that contain perforin and granzymes, and that Fas ligand may in fact be delivered to target cell Fas receptor via the same granules that deliver other granule components. This mechanism could be important for sequestering Fas ligand in the effector cell, to prevent it from causing suicide as it interacts with the receptor intracellularly. The relative importance of Fas ligand in the cytotoxic granules, as opposed to Fas ligand that is newly synthesized and directly transported to the cell surface, has not yet been fully explored. Finally, another granule protein called granulysin has been identified in human CTL, and its role in antimicrobial clearance was described above (48–50, 127). Granulysin, like perforin and the granzymes, is synthesized during T cell activation and is thought to be packaged in the same cytotoxic granule compartment. Granulysin can cause target cell membrane damage directly and can cause mitochondrial depolarization and the release of cytochrome c (47, 128). Despite causing cytochrome c release, granulysin-induced mitochondrial damage does not cause procaspase 9 activation via the classical apoptosome, but it still manages to activate caspase 3. The importance of granulysin for CTL function in vivo has been difficult to assess because the murine orthologue of this molecule has not yet been identified. Importantly, experiments with loss-of-function mice have

5 Feb 2002

12:52

AR

AR152-12.tex

AR152-12.SGM

LaTeX2e(2001/05/10)

P1: GJC

Annu. Rev. Immunol. 2002.20:323-370. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

LYMPHOCYTE-MEDIATED CYTOTOXICITY

335

strongly suggested that mice deficient for perforin have cytotoxic defects similar to those of mice deficient for granzymes A and B. This would suggest that if a murine orthologue of granulysin exists, it must cause target cell damage by a perforin-independent mechanism. Since most of the perforin-independent cytotoxicity delivered by allogeneic CTL is due to Fas (2, 129, 130), the cytotoxic role of granulysin in this setting must be limited. The cytotoxic granules are specialized lysosomes. A defect in the genes that control a crucial process in lysosomal trafficking causes the Chediak-Higashi syndrome in humans and the beige defect in mice. This trafficking regulator, called CHS1 in humans and LYST in mice, is a large cytosolic protein that is not clearly related to other known proteins (61, 131–133). This gene product affects the sorting of many lysosomal proteins in lymphocytes (and other cell types), including CTLA-4, granzymes, perforin, and MHC class II molecules in B cells (58). With the loss of function of CHS1/LYST, the granule enzymes accumulate in giant intracellular granules that apparently are unable to release the enzymes in response to appropriate stimuli, which accounts for the decreased cytotoxic activity of CTL and NK cells from patients with this syndrome.

SECRETION OF CYTOTOXIC GRANULES When an activated T cell recognizes its target, a tight junction is formed between the effector and target, and a signal is generated in the effector cell that causes its granules to migrate vectorally to the site of contact. Recently, several studies by Djeu and colleagues (134–136) have elucidated the signal transduction pathway required for granule mobilization and secretion in NK cells. These authors have shown that phosphoinositide-3 (PI-3) kinase is activated by ligation of NK cells to their targets. PI-3 kinase subsequently activates RAC-1, which activates P21activated kinase 1 (PAC-1), which in turn activates MAPK kinase and finally the extracellular signal regulated kinase (ERK). Inhibition of RAC-1 or PAC-1 mimics the suppressive activity of PI-3 kinase inhibitors, as does inhibition of MAPK-ERK kinases. Since the granule exocytosis pathway is required for NK cell cytotoxicity, inhibitors of this signal transduction pathway dramatically reduce the ability of NK cells to kill their targets. Presumably, similar signal transduction pathways are utilized to cause granule secretion in activated T cells after conjugation. Another important component of granule exocytosis has recently been discovered in association with Griscelli syndrome and the Ashen mouse (58). Mutations in RAB27A have previously been linked to immune defects in humans. Ashen mice were known to contain a mutation in RAB27A that alters its splicing and creates loss of function for this protein. Haddad et al. (60) showed that CTL and NK cells derived from Ashen mice have profoundly decreased cytotoxicity, even though they have normal Fas ligand expression and Fas-induced cytotoxicity. Ashen CTL have normal appearing granules containing normal levels of perforin and granzymes A and B, and normal polarization of granules. However, granule

5 Feb 2002

12:52

336

AR

AR152-12.tex

RUSSELL

¥

AR152-12.SGM

LaTeX2e(2001/05/10)

P1: GJC

LEY

Annu. Rev. Immunol. 2002.20:323-370. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

secretion itself is defective in CTL-derived from Ashen mice (137), even though other secretory pathways are normal (60). Patients with Griscelli syndrome, a rare autosomal recessive disease associated with hematophagocytic syndrome, also have mutations in RAB27A (59). Therefore, there is strong evidence that RAB27A is required for cytotoxic granule secretion. It is interesting to note, however, that Fas ligand–mediated killing is normal in CTLs derived from Ashen mice (60), which suggests that the Fas ligand found in granules is not required for Fas ligand– mediated cytotoxicity.

THE ESSENTIAL ROLE OF PERFORIN FOR GRANULE-MEDIATED KILLING In the original paradigm of perforin function, perforin was thought to directly induce target cell death by damaging target cell membranes, causing cell lysis similar to that induced by complement. The similarity of perforin to that of complement component 9 further suggested that this molecule was the key granule component that causes target cell death. However, the discovery that perforin-induced membrane damage was not sufficient to cause apoptosis (100), the hallmark of CTL-induced target cell death, suggested that it may instead act as a portal of entry for other cytotoxic molecules that induced the apoptotic hits. Upon release of perforin from cytotoxic granules, the molecule rapidly polymerizes in the presence of calcium to form a ring-like structure that apparently contains a central pore when it is inserted into the target cell membrane. However, the size of this pore is probably too small to permit large molecules like granzymes to enter target cells (138). Therefore, in the last several years, the idea has been questioned that perforin simply creates a channel through which other critical death-inducing enzymes pass. The insertion of polymerized perforin into target cell membranes may not be a random process, but instead may be mediated by “receptors.” This idea was first advanced by Tschopp et al. (139), who noticed that perforin and complement component 9 are distinct in their modes of target cell recognition. While complement 9 insertion is dependent on a receptor assembled from upstream complement molecules in the pathway, no components for a perforin receptor are present in the cytotoxic granules. Tschopp et al. (139) demonstrated that phosphocholine on target cell membranes acted as a specific, calcium-dependent receptor molecule for perforin. M¨uller & Tschopp (140) later described evidence for a lymphocyte membrane protein that interacted with perforin, providing a mechanism for CTL to protect themselves from inadvertent perforin membrane damage. This idea was further examined by Rochel et al. (141), who explored the role of lipid inhibition of perforin activity as a mechanism for protecting CTL from perforin-induced membrane damage. The concept of a perforin receptor has been further explored in two recent publications. Berthou et al. (142) have suggested that NK cells can release the

5 Feb 2002

12:52

AR

AR152-12.tex

AR152-12.SGM

LaTeX2e(2001/05/10)

P1: GJC

Annu. Rev. Immunol. 2002.20:323-370. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

LYMPHOCYTE-MEDIATED CYTOTOXICITY

337

lysolipid platelet activating factor (PAF), which may act as a chaperone for perforin, synergizing with it to produce membrane damage. PAF receptors were found on perforin-sensitive cell lines, but not on perforin-resistant lines; interferon-γ induced PAF receptor expression correlated with the induction of perforin sensitivity. The authors went on to propose that if interferon-γ were to fail to induce PAF receptor expression, it could make tumor cells resistant to perforin-mediated attack. A similar idea was independently explored by Lehmann et al. (143) who showed that a human leukemia cell line (ML-2) demonstrated resistance to NKmediated killing by virtue of the fact that perforin was not able to bind to the surface of these cells. They went on to examine perforin binding to leukemic cells derived from several patients and found a correlation between the failure to bind perforin and resistance to NK cell–mediated cytotoxicity. Resistance to perforin-mediated membrane damage may represent a novel mechanism of tumor cell resistance to immune killing; it certainly merits additional study.

THE TRAFFICKING OF GRANZYMES IN TARGET CELLS As noted above, the original ideas about the collaboration between perforin and granzymes for the induction of cell death have evolved considerably (105, 144). The idea that perforin created a channel through which granzymes could pass into target cells was widely held in 1996 (Figure 2). However, the channels created by polyperforin range in size up to 16 nM, which are not large enough to permit diffusion of even small proteins (i.e., 8 kDa) into cells (138). The granzymes range in size from approximately 30 to 65 kDa, and they are probably complexed with serglycin upon secretion. These observations suggested that the granzymes cannot enter target cells directly via a perforin pore. A second hypothesis for granzyme entry, namely reparative endocytosis, was offered to explain this paradox (105). In this model, perforin entry into the target cell membrane creates a signal for the target cell to repair the damage by endocytosing the perforin and surrounding plasma cell membrane; granzymes in the vicinity of the lesion are also endocytosed and ultimately delivered to the target cell cytoplasm and nucleus, where they deliver the apoptotic hits. Most of the studies described below utilize in vitro systems where target cells are bathed in solutions of perforin and granzymes. It is not yet known whether the models provided by these studies are physiologically relevant; in vivo, cytotoxic lymphocytes secrete large amounts of granule proteins onto a very small patch of target cell membrane defined by the conjugation site. These conditions cannot currently be reproduced in vitro, but they may be very important for defining perforin pore size and/or mechanisms of granzyme entry in vivo. Several groups have now demonstrated that purified granzyme B can enter target cells without any chaperones or perforin (145–147). These studies demonstrated that the internalized granzyme B requires an additional signal to cause cell death (that can be provided by either perforin, adenovirus, or bacterial toxins)

5 Feb 2002

12:52

Annu. Rev. Immunol. 2002.20:323-370. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

338

AR

AR152-12.tex

RUSSELL

¥

AR152-12.SGM

LaTeX2e(2001/05/10)

P1: GJC

LEY

Figure 2 Mechanisms of granzyme entry into target cells. The effector cell is shown at the top, and the target cell on the bottom. On the left side, the traditionally held view of perforin as a pore-forming molecule is shown. In this model, perforin secreted by the effector cell polymerizes in the presence of calcium and forms a channel through which the granzymes and other constituent granule proteins pass into the target cells. On the right side of the panel, a more recent model for granzyme entry is presented. In this model, perforin enters the target cell membrane and creates a stimulus for repair. Perforin enters the target cell via reparative endocytosis. In one model of granule entry, the granzymes would be inadvertently taken into the target cell with the perforin during this process. In a more recent model, granzyme B enters the target cell by virtue of its binding to the cation-independent mannose-6-P receptor (CI-MPR). After the receptor is internalized, the granzyme B is released into an endolysosomal compartment where it is rendered harmless. Internalized perforin provides a signal for granzyme B to leave this compartment, where it can then cleave cytoplasmic and nuclear apoptotic substrates. See text for details.

5 Feb 2002

12:52

AR

AR152-12.tex

AR152-12.SGM

LaTeX2e(2001/05/10)

P1: GJC

Annu. Rev. Immunol. 2002.20:323-370. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

LYMPHOCYTE-MEDIATED CYTOTOXICITY

339

(138). This second signal is responsible for trafficking the intracellular granzyme B from an endolysosomal compartment (where it is harmless to the cell) into the nucleus. Granzyme B delivered into target cells by microinjection can directly cause apoptosis (147), suggesting that nuclear entry of granzyme B (148–150) does not require a second signal. Similar results have been described for granzyme A (151). Even though these studies clearly demonstrate that granzymes can enter target cells autonomously, they do not demonstrate whether this pathway is minor or major for CTL-mediated granzyme delivery. The notion that granzymes could enter target cells independently of perforin led to a search for granzyme receptors, which were first described by Motyka et al. (152). Granzymes A and B were previously known to be targeted to the cytotoxic granules via the mannose-6 phosphate receptor (MPR) (118). Motyka et al. therefore explored the possibility that granzyme B could enter target cells via the MPR, and they discovered that native granzyme B binds to both the cationindependent (CI) and cation-dependent (CD) MPR (Figure 2). However, they found that only the CI-MPR (also known as the insulin-like growth factor 2 receptor) was required for the induction of granzyme B–mediated apoptosis in an in vitro system. In addition, the expression of the CI-MPR was required for the killing of allogeneic cells in an in vivo model system. Clearly, these results suggest that granzyme entry into target cells could be modified by CI-MPR receptor blockade. However, the relative utilization of this pathway for granzyme entry into target cells has not yet been clearly defined. Does a large percentage of the granzyme B delivered by a CTL enter the target cell via this mechanism, or is it only a small fraction? Importantly, active recombinant granzyme A and B have been produced in systems (yeast and baculovirus) that do not attach Man-6-P to proteins. Recombinant granzyme B is active against peptide and protein substrates (116, 153) and can cause the induction of apoptosis when it is delivered to target cells in the presence of perforin (124). However, the relative apoptotic activities of Man-6-P modified granzymes and their recombinant counterparts have not yet been reported.

THE STRUCTURE AND FUNCTION OF GRANZYME B Granzyme B Substrate Specificity Granzyme B, a serine protease, was originally defined as an aspase because of its preference to cleave after aspartic acid in the P1 position (108). In this regard, it is similar to the caspases, which also prefer to cleave with aspartic acid at P1. However, purified human granzyme B has a unique specificity, different in the P2, P3, and P4 positions, so that the preferred recognition motif for granzyme B is (I/V) EPD (154). The substrate specificity for recombinant rat granzyme B was further characterized by Harris et al. (155), who discovered that the enzyme has an extended substrate specificity, which includes any amino acid in the P10 position and Gly in the P20 position. The activity of the enzyme was highly dependent on the length and sequence of substrate peptides, implying that the targets for this

5 Feb 2002

12:52

340

AR

AR152-12.tex

RUSSELL

¥

AR152-12.SGM

LaTeX2e(2001/05/10)

P1: GJC

LEY

enzyme are highly restricted. The extended substrate specificity correlated well with the presence of granzyme B cleavage sites in many of the caspases and other substrates known to be cleaved by granzyme B upon target cell entry.

Annu. Rev. Immunol. 2002.20:323-370. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

Granzyme B Structure The crystal structures of both rat and human granzyme B have recently been elucidated (156–158). The structure of rat granzyme B in complex with a macromolecular inhibitor revealed the structural determinants that explain its extended substrate specificity. The primary specificity for Asp in the P1 position occurred through a buried side chain of arginine at position 226. Nine additional amino acids were shown to make contact with the substrate, and they defined the extended substrate specificity profile. The structure of the human enzyme revealed that the S1 subsite of this enzyme was larger and less charged than the corresponding Asp-specific site of the caspases or of rat granzyme B. These differences are significant enough to be relevant for substrate specificity, and they point out the importance of species of origin when granzymes are being compared. Subtle differences in peptide substrate specificities have been detected for rat, mouse, and human granzyme B (N. Thornberry, personal communication), and they will need to be taken into account as inhibitors are prepared, as cellular substrates are examined, and as purified or recombinant enzymes are used in the killing of cells of a different species.

The Death Pathways of Granzyme B After granzyme B is released from its endolysosomal compartment, it is apparently rapidly trafficked to the nucleus by a mechanism that does not require perforin or alterations in nuclear pore size (148–150). Presumably, free granzyme B in the target cell initiates cell death by cleaving a variety of protein substrates that are either directly or indirectly linked to the induction of DNA fragmentation and cell death (Figure 3). Over the past several years, many studies have clearly shown that a number of procaspases (including caspases 2, 3, 7, 8, 9, and 10) are substrates of granzyme B both in vitro and in vivo (109, 159–163). A large number of studies have also evaluated the requirement of caspases in the target cell for granzyme B’s ability to induce target cell apoptosis (111, 124, 162, 164–169). These results have been somewhat variable from group to group, but ultimately, a fairly clear consensus has emerged. Even though granzyme B can clearly induce target cell DNA fragmentation and apoptosis in cells that contain broad spectrum inhibitors of caspases, or that lack caspase 3 (124), the induction of death occurs more rapidly in cells that contain active caspases, suggesting that they amplify or feed forward the granzyme B death signal. The ability of granzyme B to cause DNA fragmentation in cells that lack functional caspases suggested that granzyme B could directly activate apoptotic nucleases. Wang & Nagata and their colleagues (170–173) independently identified such a nuclease, known as caspase-activated DNAse (CAD, also known as DFF40). CAD exists as a heterodimer with ICAD (Inhibitor of CAD, also known as DFF45)

5 Feb 2002

12:52

AR

AR152-12.tex

AR152-12.SGM

LaTeX2e(2001/05/10)

P1: GJC

Annu. Rev. Immunol. 2002.20:323-370. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

LYMPHOCYTE-MEDIATED CYTOTOXICITY

341

Figure 3 Currently identified pathways of granzyme B action. After granzyme B enters target cells, it cleaves and activates procaspase 3 (and several other procaspases), Bid, and ICAD. The granzyme B proapoptotic signal is amplified by caspase activation and/or Bid cleavage and translocation to the mitochondrial membrane, where Bax and Bak then form a channel that permits cytochrome c release. Granzyme B is also capable of directly acting on mitochondrial membranes to cause depolarization in the absence of cytochrome c release. “X” refers to a putative cellular cofactor for granzyme B-induced mitochondrial depolarization (186). All of these pathways initiate either DNA fragmentation or apoptosis in the target cell.

in the nucleus, where CAD is retained in its inactive form. After a cell receives a death signal, caspase 3 is activated by one of a variety of pathways, and it can then cleave ICAD/DFF40, a process that facilitates the assembly of CAD into its active form (170, 174). Thomas et al. (124) showed that ICAD is a direct substrate of granzyme B and that CAD can be directly activated in vitro and in target cells by granzyme B. Furthermore, ICAD is not cleaved during the induction of apoptosis by granzyme B–deficient CTL. Finally, ICAD-deficient cells (175, 176) are partially resistant to granzyme B–mediated killing, which demonstrated the biological significance of the ICAD pathway for granzyme B-induced death (124). Similar studies by Sharif-Askari et al. (177) confirmed these results. However, since a large

5 Feb 2002

12:52

Annu. Rev. Immunol. 2002.20:323-370. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

342

AR

AR152-12.tex

RUSSELL

¥

AR152-12.SGM

LaTeX2e(2001/05/10)

P1: GJC

LEY

percentage of cells that contain neither ICAD nor functional caspases are susceptible to granzyme B–mediated death, additional death pathways must also exist. A number of observations suggested that one of these pathways may involve mitochondria, since Bcl-2 variably blocks death caused by perforin and granzymes (178–183). Barry et al. (167), Sutton et al. (169), Heibein et al. (184), and Alimonti et al. (185) all went on to show that granzyme B is capable of directly cleaving Bid upon target cell entry. Bid cleavage resulted in translocation of tBid to the mitochondria, where it interacted with its receptors Bax and Bak to cause cytochrome c release. Cytochrome c then activates the apoptosome, which activates procaspase 9, and ultimately caspase 3 (Figure 3). Although granzyme B can clearly cleave and activate Bid, and cause cytochrome c release, Thomas et al. (186) have recently shown that neither Bid-deficient nor Bax/Bak doubly deficient mouse embryo fibroblasts are resistant to granzyme B–induced death. Granzyme B can cause mitochondrial depolarization even in the absence of cytochrome c release, a process that requires neither the permeability transition pore or the Bax/Bak receptors. These data all suggest that granzyme B amplifies its death signal upon target cell entry by cleaving Bid and causing mitochondrial dysfunction, but it also suggests that these pathways are not required for induction of death. Commitment to death probably proceeds more slowly in the absence of Bid, but granzyme B can persistently attack alternative targets, like ICAD, to cause the ultimate death of the cell. Granzyme B also cleaves a variety of other substrates upon target cell entry, but the relationship of these events to the induction of apoptosis is not yet clear. For example, granzyme B cleaves poly (ADP-ribose) polymerase (PARP) (187), DNA-PKcs and NuMA (188), Filimin (189), cartilage proteoglycan (190), nuclear lamins (191), and a variety of autoantigens (192) during the induction of target cell death. The significance of these cleavage events for cellular homeostasis and/or the production of autoimmune diseases are not yet clear.

GRANZYME A Granzyme A was the first serine proteinase activity to be discovered in cytotoxic granules (193) but far less is known about its mechanism of action in inducing cell death than that of granzyme B. The structure of granzyme A has not yet been solved. Granzyme A is a tryptase and prefers to cleave synthetic substrates with Arg or Lys at the P1 position (108). Granzyme A exists in granules as a disulfidelinked homodimer, which is probably complexed with serglycin, like granzyme B (123). After granzyme A is released from activated CTL, it can circulate in an active form, bound to proteoglycans that protect it from inactivation by proteins like anti-thrombin III or α2-macroglobulin (194). Shi et al. (101) and Shiver et al. (103) independently showed that granzyme A is a pro-apoptotic enzyme, using different approaches. When delivered to target cells with perforin, purified granzyme A can clearly induce target cell apoptosis (101).

5 Feb 2002

12:52

AR

AR152-12.tex

AR152-12.SGM

LaTeX2e(2001/05/10)

P1: GJC

Annu. Rev. Immunol. 2002.20:323-370. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

LYMPHOCYTE-MEDIATED CYTOTOXICITY

343

When cotransfected with perforin into rat basophilic leukemia cells, granzyme A can cause target cell death with DNA breakdown (103). The timing of cell death induced by granzyme A, however, is much slower than that of granzyme B; granzyme A requires many hours to induce apoptosis within in vitro–reconstituted systems. The reasons why granzyme A induces death so slowly are unknown. They could include: (a) a proteolytic cascade that involves a large number of enzymes and substrates, and/or (b) a very slow catalytic rate for a rate-limiting step in death induction, or (c) a specialized compartment of granzyme A–expressing CTL that delivers its death signals more slowly than the cells that deliver granzyme B. None of these potential mechanisms has yet been tested. The completely different specificities of granzymes A and B suggested either that these enzymes could be synergistic or that they may act through nonoverlapping, alternative pathways as fail-safe mechanisms for the induction of target cell death. Although some evidence has been gathered for a synergistic interaction of granzymes A and B (104), experiments from knockout mice (195) strongly suggest that these pathways are independent. The protein targets of granzyme A have been extensively investigated. Granzyme A is known to bind to and cleave nucleolin (196), activate the thrombin receptor on neuronal cells and astrocytes (197), cleave and activate interleukin-1β (198), and cleave PHAP II, a ubiquitous, putative HLA-associated protein (199). Granzyme A binds to heat shock protein 27 in target cells but does not cleave it (200), and it enhances DNA accessibility to exogenous endonucleases by degrading histone-H1 (201). Finally, granzyme A directly cleaves lamins, thereby disrupting the nuclear lamina and causing nuclear breakdown during the induction of apoptosis (191). Some of these substrates are highly likely to be involved in granzyme A–mediated death, but definitive experiments to assign the requirement of each of these substrates for cell death have not yet been performed. The cellular pathways used by granzyme A to induce death appear to be independent from those utilized by granzyme B. Granzyme A does not appear to activate caspases in target cells (202, 203), nor does it induce cleavage of many granzyme B substrates, like caspase-3, DNA-PK, or PARP. Furthermore, granzyme A–induced death does not cause oligonucleosomal DNA degradation, but rather it appears to cause single-stranded DNA breaks, suggesting that it activates a different nuclease than granzyme B or the caspases (203). The uniqueness and alternative features for this pathway were further demonstrated by studies of the granzyme A knockout mouse (see below).

ORPHAN GRANZYMES We refer to all of the granzymes of unknown function as the orphans. A catalog of these orphan enzymes and their specificities has recently been assembled by Kam et al. (204); they include granzymes C, D, E, F, G, and K in the mouse (205), granzymes H, K, and M in the human, and granzymes C, I, J, K, and M in the rat. The

5 Feb 2002

12:52

Annu. Rev. Immunol. 2002.20:323-370. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

344

AR

AR152-12.tex

RUSSELL

¥

AR152-12.SGM

LaTeX2e(2001/05/10)

P1: GJC

LEY

specificities of these enzymes are largely unknown, except for human granzyme H, which is a chymase (206), granzyme K, which is a tryptase, and granzyme M, which is a metase (204). The roles of these enzymes for CTL-mediated death are all essentially unknown at this time. The granule genes are organized into three independent clusters in the genome. Granzymes A and K are tightly linked on chromosome 13 in the mouse (207) and on chromosome 5 in humans. Granzymes B, C, D, E, F, and G are tightly linked in a gene cluster that also includes mouse cathepsin G and several mast cell chymases on chromosome 14 (208). Human granzymes B and H are tightly linked with the human cathepsin G and mast cell chymase gene on chromosome 14 (209, 210). Finally, human Metase is linked to the neutrophil elastase, proteinase-3, and azurocidin genes on chromosome 19 pTer (211, 212). In vitro, granzymes A and B are highly expressed, along with perforin, in activated T cells and NK cells. In human NK cell lines, granzyme H is also expressed, but at levels considerably lower than granzyme B (210). In the mouse, granzymes A and B are abundantly expressed in cytotoxic T cells and NK cells; granzymes C, D, F, and K are expressed at high levels in NK cells and LAK cell preparations (129, 208). However, the pattern of expression of these orphan granzymes in CTL activated in vivo has not yet been reported; the biologic significance of these restricted patterns of expression from in vitro activations is therefore unknown. Human granzyme H is a chymase that is predominantly expressed in the NK cell compartment; the gene that encodes it lies between granzyme B and cathepsin G (210). Murine granzymes C-G are expressed in in vitro–generated LAK and NK cells, and their location between granzyme B and cathepsin G suggests that they may serve similar functions in the mouse and the human. Indeed, human granzyme H 50 flanking sequences can target the expression of SV40 T-antigen to murine NK-T progenitors and activated LAK cells (213). All of these data suggest that the murine granzymes downstream from granzyme B may indeed subserve functional roles for murine cytotoxic lymphocytes. Although there is no data to support this hypothesis as yet, it is tempting to speculate that mice may have evolved a more divergent set of granzymes to protect them from a more diverse set of microbial pathogens than that encountered by humans (it seems unlikely that selection pressure was provided by tumors, because the reproductive cycle of mice is very short, and the life span of feral mice is measured in months). Regardless, the broad array of granzymes expressed in CTL, and the different specificities of these enzymes, suggest that they may play specialized roles in the induction of target cell death in specific physiologic situations.

LESSONS FROM THE KNOCKOUTS Perforin Deficiency in Mice and Humans Four independent groups created perforin-deficient mice (5, 214–216), and the phenotypes of these mice have been extensively described here and elsewhere (1). Because perforin is required for generating trafficking signals for granzymes

5 Feb 2002

12:52

AR

AR152-12.tex

AR152-12.SGM

LaTeX2e(2001/05/10)

P1: GJC

Annu. Rev. Immunol. 2002.20:323-370. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

LYMPHOCYTE-MEDIATED CYTOTOXICITY

345

after they are delivered to target cells, perforin deficiency causes functional pangranzyme deficiency. CD8+ T cells and NK cells are dramatically altered in their ability to cause the membrane changes associated with the induction of target cell apoptosis. Perforin deficiency results in increased susceptibility to a variety of viral infections and is associated with delayed tumor clearance in a variety of whole animal systems (1). It is also associated with the mitigation of GvHD (GvHD) severity in some model systems, especially those that are heavily dependent on the CD8+ compartment for disease (217–224). However, a role for the perforin pathway can be detected in graft vs. host models that are dependent on CD4+ cells as well (25, 225, 226). CD8+-dependent engraftment also requires perforin (227). Although perforin-deficient mice were initially thought to have no defects in lymphocyte homeostasis (1), a series of subsequent studies has suggested otherwise (112). In some model systems, perforin is involved in downregulating peripheral T cells after activation (65, 228). Activated CD8+ cells accumulate in perforindeficient mice after LCMV infection (62, 65), and perforin x Fas ligand double deficient mice die early of pancreatitis associated with infiltration of macrophages, MAC-1+ T cells, and CD8+ cells in the pancreas (229). Importantly, these defects in lymphocyte homeostasis do not occur spontaneously (as they do in FasL or FasR deficient mice); a second mutation or an infection is required to elicit the phenotype. Recently, Stepp et al. (230) described homozygous loss-of-function defects in the perforin gene in human patients with familial hemophagocytic lymphohistiocytosis (FHL). This disease is characterized by uncontrolled activation of T cells and macrophages and overproduction of inflammatory cytokines. FHL mapped to 10q21-22, the location of the perforin gene, in a fraction of families with this disorder. A more extensive analysis of additional families revealed perforin mutations in about 20% of all FHL patients investigated; the nature of the mutations for the other patients is not yet known (231). Interestingly, mice deficient for perforin do not spontaneously develop anything that resembles FHL. The addition of a viral infection or a second defect in a pathway that also controls lymphocyte homeostasis (i.e., Fas) is required for the development of the hemophagocytic syndrome. The nature of this striking difference between perforin-deficient humans and mice is not yet clear.

Granzyme B Cluster Deficiency Heusel et al. (106) first described a mouse with a targeted null mutation of the granzyme B gene; the mutant allele contained a PGK-neo cassette in the granzyme B gene. This mouse was null for granzyme B expression in CTL, but on careful analysis, it was also profoundly deficient for the expression of granzymes C, D, and F in the LAK cell compartment, suggesting that the PGK-neo cassette located at the 50 end of the granzyme gene cluster caused downregulation of all the genes in the domain (208). All of the published results for granzyme B deficiency have utilized these mice, and therefore all of the defects described could be due to granzyme B and/or one or more of the orphan genes that lies downstream. Recently, Thomas

5 Feb 2002

12:52

Annu. Rev. Immunol. 2002.20:323-370. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

346

AR

AR152-12.tex

RUSSELL

¥

AR152-12.SGM

LaTeX2e(2001/05/10)

P1: GJC

LEY

et al. (D. Thomas, R. Behl, T. J. Ley, unpublished data) have removed the PGK-neo cassette from the granzyme B locus using LoxP-Cre technology and showed that the in vitro cytotoxicity defect present for the granzyme B cluster–deficient CTL is identical for CTL deficient for granzyme B only. Therefore, granzyme B itself clearly plays a major role for the in vitro phenotype previously described. In vivo experiments comparing these mice are in progress. Granzyme B–deficient CTL are profoundly deficient in their ability to induce DNA fragmentation in target cells, even though perforin-induced membrane damage caused by these CTL is normal. Prolonged incubation of granzyme B–deficient effectors with target cells results in restoration of nearly all of the cytotoxic potential of the effectors (107, 130). Several groups showed that the perforinindependent component of this cytotoxicity could be accounted for the by the Fas pathway (2, 129, 130); the second component is a perforin-dependent mechanism, which is due to the activity of granzyme A (see below). The early pathway of cytotoxicity that requires granzyme B is important in vivo. Effectors deficient for granzyme B have an attenuated acute GvHD phenotype in partially and fully allogeneic model systems (232). This attenuation is not nearly as prominent as that of perforin deficiency, since granzyme B represents only one of the granzymes delivered to target cells by perforin. Granzyme B–deficient mice have not yet been systematically challenged with a variety of viruses because a minimal phenotype is predicted, due to the fact that these mice contain alternative pathways for clearing viruses. Granzyme B–deficient humans have not yet been described.

Granzyme A Deficiency Ebnet et al. (233) and Shresta et al. (207) both generated granzyme A–deficient mice. Granzyme A lies in a cluster that also contains granzyme K in the mouse. Retention of the PGK-neo cassette in granzyme A knockout mice did not affect granzyme K expression, and these mice do indeed have residual tryptase activity in their CTL (207, 234). Granzyme A–deficient mice exhibited no detectable abnormality of cytotoxicity in vitro, probably because these effector cells had normal amounts of perforin and granzyme B. However, the residual tryptase activity provided by granzyme K in these mice may also partially account for the minimal phenotype. Granzyme A–deficient mice fail to clear the poxvirus ectromelia as efficiently as wild-type mice (36, 235, 236). Ectromelia contains within its genome a CrmA orthologue, a serpin that may be capable of inactivating granzyme B. With this infection, therefore, granzyme B may be inactivated by the virus-produced CrmA, so that granzyme A–deficient cells exhibit the delayed-clearance phenotype. Granzyme A–deficient mice also fail to restrict the spread of herpes simplex virus in the peripheral nervous system of mice (237), but the mechanism of this effect is not yet clear. Granzyme A–deficient humans have not yet been described.

5 Feb 2002

12:52

AR

AR152-12.tex

AR152-12.SGM

LaTeX2e(2001/05/10)

P1: GJC

LYMPHOCYTE-MEDIATED CYTOTOXICITY

347

Annu. Rev. Immunol. 2002.20:323-370. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

Granzyme A × B Cluster Deficient Mice The granzyme B cluster–deficient mouse was intercrossed with the granzyme A mouse generated by Ebnet et al. (233) and also with the one generated by Shresta et al. (207); these two phenotypes were similar (195, 238). Granzyme A and B doubly deficient CTL can release chromium from target cells, and they have normal abundance and activity of perforin. Regardless, cytotoxic effectors from these mice had an apoptotic defect that was virtually equivalent to perforin-deficient effectors at all time points tested, strongly suggesting that granzyme A represents the perforin-dependent, late pathway in granzyme B-deficient mice. In an allogeneic bone marrow–transplant model that was dependent on CD8+ T cells, granzyme A × B cluster–deficient mice had a markedly attenuated incidence of acute GvHD (that was virtually equivalent to that of perforin-deficient mice) (195). That result suggested that perforin itself does not cause target cell death in this in vivo system and that the granzyme A and B cluster enzymes provided the lethal hits. Granzyme A × B double knockout mice have also been evaluated in a viral clearance model. M¨ullbacher et al. showed that granzyme A × B cluster-deficient mice were as susceptible to ectromelia as perforin-deficient mice, again suggesting that perforin itself was not the direct cause of death of the ectromelia-infected cells; the lethal hits are provided by granzyme A and B cluster enzymes. However, Davis et al. (239) have recently suggested that granzyme A × B cluster-deficient mice resisted RMAS tumor cells (cleared by NK cells) as efficiently as wild-type mice, while perforin-deficient mice are highly susceptible to these tumors. Perforin-deficient mice were also more susceptible to tumor initiation by methylcholanthrene than were granzyme A × B mice. These results suggest that perforin itself, or some other granule component delivered by perforin, is most critical for the in vivo anti-tumor effector functions of NK cells and cytotoxic T cells. In vitro, the defect in the killing of many tumor cells lines has been compared for perforin-deficient vs. granzyme A × B cluster-deficient mice, and this difference has not been appreciated. These results have not yet been reproduced by other laboratories, but they are clearly provocative and require additional experimentation with other model systems.

DPPI (Cathepsin C) Deficiency in Mice and Humans The study of DPPI-deficient mice (117) confirmed that this enzyme is responsible for the activation of granzymes A and B in CTL in vivo. As expected, DPPIdeficient CTL had an in vitro cytotoxicity defect that was virtually equivalent to that of mice deficient for granzymes A and B. Virtually all of granzymes A and B produced in DPPI-deficient CTL had retention of the N-terminal dipeptide or abnormal processing of the N terminus of the molecule, and no associated enzymatic activity. Recent studies have clearly shown that DPPI is also required for the processing and activation of mast cell chymases, (but not tryptases) (240), and it is responsible for activation of murine neutrophil elastase, cathepsin G,

5 Feb 2002

12:52

Annu. Rev. Immunol. 2002.20:323-370. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

348

AR

AR152-12.tex

RUSSELL

¥

AR152-12.SGM

LaTeX2e(2001/05/10)

P1: GJC

LEY

and proteinase 3 (117, 241). Therefore, DPPI-deficient mice have broad defects in serine proteinase activities in several hematopoietic lineages. Shortly after the description of DPPI-deficient mice, human patients with lossof-function DPPI mutations were described by Toomes et al. (242) and Hart et al. (243); these patients had Papillon-Lef`evre syndrome, also known as keratosis palmoplanterus with periodontopathia. This autosomal recessive disorder is associated with premature tooth loss due to periodontal disease, and with thickening of the skin. Homozygous loss-of-function mutations for DPPI/cathepsin C were identified in several consanguineous Papillon-Lef`evre kindreds (244) and also for additional kindreds with the very similar Haim-Munk syndrome (245) and prepuberal periodontitis (246). The natural history of these patients is not well understood at this time (i.e., do they have excessive viral infections or high cancer incidence?). Furthermore, cytotoxic lymphocytes from these individuals have not yet been examined to see whether they contain active granzymes. Based on the hypothesis that the granule exocytosis pathway is essential for lymphocyte homeostasis, these patients would be expected to have an hematophagocytic syndrome if their granzymes are nonfunctional. However, descriptions of these patients from the literature suggests that this is not the case (247). DPPI-deficient mice do not reproducibly develop skin thickening. Since mouse dentition is fundamentally different from human dentition, no periodontal disease has been observed in these animals. Importantly, DPPI-deficient mice can process some granzyme C to its fully mature form, suggesting that mice contain at least one other enzyme that is capable of processing some of the orphan granzymes (117). Whether humans possess this activity is not known. Mice and humans with DPPI deficiency therefore have somewhat different phenotypes, with the effect of the mouse loss-of-function mutation again being milder than that of the human mutations. Again, however, it remains to be seen whether the cytotoxic lymphocytes of human DPPI–deficient patients lack functional granzymes A and B.

ARE GRAFT VS. HOST AND GRAFT VS. LEUKEMIA (GvL) EFFECTS MEDIATED BY THE SAME OR DIFFERENT CYTOTOXIC MECHANISMS? It has been clear for many years that much of the benefit enjoyed by leukemia patients after allogeneic transplantation is a potent GvL effect that is provided by the T cells in the donor graft. This potential has been further exploited in recent years using donor lymphocyte infusions (DLI), which have been used successfully in patients with chronic myeloid leukemia in relapse after allogeneic stem cell transplantation. A small dose of donor T cells can cause potent GvL effects and clear the residual leukemia clone, but these T cells usually cause the development of GvHD. It is crucial, therefore, to understand whether the graft vs. host and GvL effects can be separated. The question, simply put, is whether the granule

5 Feb 2002

12:52

AR

AR152-12.tex

AR152-12.SGM

LaTeX2e(2001/05/10)

P1: GJC

Annu. Rev. Immunol. 2002.20:323-370. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

LYMPHOCYTE-MEDIATED CYTOTOXICITY

349

exocytosis pathway or the Fas pathway contributes more heavily to one or the other. Three recent studies, summarized below, have addressed this issue. Before reviewing these studies, some background information is required. Since the development of perforin-deficient and Fas ligand–deficient mice in the mid1990s, a large number of bone marrow transplantation experiments have been performed that evaluated each of these effector pathways for their contributions to GvHD in a variety of model systems. These systems have many variables that differ from study to study, including strain pairs, the dose and purity of T cells used in the donor product, the dose of bone marrow cells and purity of hematopoietic stem cells in that product, the conditioning regimen, the housing and feeding conditions, and even the evaluation of GvHD severity. In each study, these variables are different, and although controlled for in each study, it is very difficult to compare two studies of this kind and draw absolute conclusions. Nonetheless, the overall results support the notion that both the Fas and perforin pathways contribute significantly to acute GvHD in some settings (25, 217–219, 248). The relative contribution seems to be determined by the model system. Importantly, few (if any) of the mouse transplantation systems used to evaluate GvHD highly resemble chronic human GvHD, which remains the rate-limiting obstacle for human transplantation. Human chronic GvHD is usually caused by minor histocompatibility differences, and it develops weeks to months after transplantation occurs. Tsukuda et al. (222) evaluated leukemia development and GvHD in lethally irradiated F1 recipient mice (B6 × DBA/2) transplanted with bone marrow and spleen cells from C57 Bl/6 donors (WT, Fas ligand-deficient, or perforin-deficient). Just before transplant, the mice were inoculated with L1210 leukemia cells or P815 mastocytoma cells, and these mice were compared with nonleukemic controls. Deficiency of either perforin or Fas in this model system reduced lethal GvHD, but the GvL effect was apparent only in the mice that received their grafts from wild-type mice or Fas ligand–deficient mice; the use of perforin-deficient cells for donor engraftment resulted in early death from leukemia. These results suggested that loss of the Fas pathway reduced GvHD without impairing the GvL effect, which was perforin dependent. However, these results could have been influenced by the fact that both L1210 and P815 cells express class I but not class II. Pan et al. (223) used a model system that involved G-CSF mobilized allogeneic peripheral stem cells, which cause a reduced severity of acute GvHD in murine models, possibly because of immunomodulation of cells in the donor graft (i.e., enhanced mobilization of Tc2 cells over Tc1 cells). These authors determined that G-CSFmobilized allogeneic stem cell products caused less GvHD, but maintained potent GvL effects that were perforin dependent. The model system used was similar to that of the previous study (B6→B6D2F1), and leukemia was again induced with P815 mastocytoma cells (H-2d). Finally, Schmaltz et al. (224), using a parent→F1 model system like that previously described (and leukemia induction with P815 mastocytoma cells or 32DP210 cells), similarly concluded that GvHD could be mitigated by Fas ligand deficiency without an increase in the incidence of leukemia, but that perforin deficiency caused mice to prematurely develop leukemia.

5 Feb 2002

12:52

Annu. Rev. Immunol. 2002.20:323-370. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

350

AR

AR152-12.tex

RUSSELL

¥

AR152-12.SGM

LaTeX2e(2001/05/10)

P1: GJC

LEY

All of these studies concluded that the allogeneic GvL effect was predominantly due to the perforin pathway. Even though the perforin and Fas systems both contributed to the induction of acute GvHD in these experimental models, inhibition of only the perforin pathway was detrimental to the outcome of allogeneic transplantation that requires GvL for curative intent. In a parallel study that examined the cytotoxic mechanisms essential for killing a syngeneic myeloid leukemia cell line that is a target for both CD4+ and CD8+ cells, Hsieh & Korngold (249) again found that the perforin pathway was required. The perforin pathway may therefore be required for both GvL effects and antileukemia surveillance. Similarly, Smyth et al. (249a) have recently shown that perforin-deficient mice have a high incidence of spontaneous lymphomas (of T, B, or NK lineages) that develop after long latent periods, suggesting that the perforin pathway is indeed important for the immune surveillance of some tumor types.

GRANZYME INHIBITORS For many years, investigators have questioned how CTL can avoid suicide after granule exocytosis. Presumably, some granule contents must inadvertently re-enter the effector cell, which should cause effector cell death. Differential insertion of perforin into target cell membranes vs. effector cell membranes had been postulated previously. More recently, Bird and his colleagues (250–253) have provided a series of observations that suggest that effector cells contain a potent inhibitor of granzyme B, known as proteinase inhibitor 9 (PI-9). This human serpin is found in both the cytoplasm and nucleus of CTL, and it can form a tight complex with granzyme B via a classical serpin-proteinase interaction. When PI-9 is overexpressed in target cells, it can inhibit the induction of apoptosis either by whole CTL or by the addition of purified perforin and granzyme B. PI-9 does not inhibit most caspases, strongly suggesting that the inhibition of granzyme B alone is sufficient to block apoptosis even by whole CTL (which can kill either by perforin or Fas-based mechanisms). The true importance of this molecule for protecting CTL against accidental suicide has not yet been evaluated because the true murine orthologue of PI-9 has not yet been identified. A murine homologue, SPI-6 (with 68% identity to PI-9), and at least six other highly related murine serpins exist, but the true orthologue remains to be defined (P. Bird, personal communication). The studies of Bird and colleagues have provided strong evidence that CTL take significant precautions to protect themselves from granzyme B that is misdirected to the effector cell after granule secretion. The overexpression of PI-9 or its close relatives in tumor cells or virus-infected cells might be expected to protect these cells from death as well. Indeed, Madema et al. (253a) have recently shown that SPI-6 and PI-9 are expressed in a variety of murine and human tumors, respectively, and that overexpression of SPI-6 in tumor cells can protect them from CTLmediated clearance in vivo. However, the relative importance of this mechanism for tumor “escape” from immune surveillance remains to be defined.

5 Feb 2002

12:52

AR

AR152-12.tex

AR152-12.SGM

LaTeX2e(2001/05/10)

P1: GJC

LYMPHOCYTE-MEDIATED CYTOTOXICITY

351

Annu. Rev. Immunol. 2002.20:323-370. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

Similar studies for inhibitors of granzyme A and effector cells have not yet been reported. However, two molecules are known to interact with granzyme A, antithrombin III and α-2 macroglobulin. Both of these molecules can stably interact with granzyme A and neutralize it; complexes of the inhibitors and granzyme A have been detected in the circulation (194). It is not clear whether these molecules are present in activated CTL or NK cells, or whether they can inhibit granzyme A-induced death.

VIRUS-ENCODED INHIBITORS OF GRANZYME B If viruses are targeted for immune destruction by CD8+ T cells, it would be highly likely that the virus-infected cells could survive for longer periods of time if they were able to inhibit the action of perforin or granzyme B, based on the lossof-function models presented above. It does not seem likely that viruses would gain anything by creating inhibitors to granzyme A. However, it is possible that specific granzymes, especially C-G in the mouse, may have roles that are restricted to specific classes of viruses, although there are no data as yet to support this hypothesis. Accordingly, several different viruses have now been reported to encode inhibitors of granzyme B, and several different mechanisms of granzyme B inhibition have been implicated by these viral processes. The first, and best described, is that of the poxvirus-encoded cytokine response modifier A gene (CrmA). Quan et al. (254) showed that CrmA does associate with granzyme B in vitro and inhibit it; Tewari (255) showed that overexpression of CrmA in target cells could inhibit CTL-mediated killing, but predominantly through a Fas-mediated pathway. Macen et al. (256) also suggested that CrmA (Spi2) could inhibit both Fas- and granzyme B-mediated killing. Subsequently, Zhou et al. (257) indeed showed that CrmA was capable of interacting with several caspases and inhibiting them; the most prominent inhibition was to caspase 1 and caspase 8. The presence of CrmA in a poxvirus-infected cell therefore would be expected to reduce CTL-mediated killing by inhibiting both granzyme B and caspases activated by the Fas pathway. A second described mechanism of granzyme B inhibition is a loss of granzyme B mRNA in virus-infected T cells. Parainfluenza virus type 3 can directly reduce granzyme B mRNA abundance in virus-infected CTL (258). The mechanism of granzyme B mRNA loss caused by the virus was not elucidated (i.e., transcriptional downregulation vs. increased RNA turnover), but it was specific for granzyme B (and not granzyme A) mRNA. No additional reports of direct effects of viruses on the RNA abundance of cytotoxic granule genes have subsequently been reported. Finally, Andrade et al. (259) have reported that the adenovirus assembly protein L4-100K is a granzyme B substrate and that it also potently inhibits the enzyme. The inhibition is dependent on the presence of specific aspartic acid residue within this target protein, found within a classical granzyme B consensus motif. Interestingly, the 100K protein inhibited granzyme B specifically and had no activity

5 Feb 2002

12:52

Annu. Rev. Immunol. 2002.20:323-370. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

352

AR

AR152-12.tex

RUSSELL

¥

AR152-12.SGM

LaTeX2e(2001/05/10)

P1: GJC

LEY

against caspases 1-13 (in contrast to CrmA, which interacts with both granzyme B and several caspases). This is therefore the first description of a viral inhibitor of granzyme B that specifically targets this enzyme by acting as a decoy substrate. The idea of specifically targeting granzyme B, as opposed to a broader, pan-caspase inhibition, is supported by the many studies described above that have made it clear that granzyme B does not require caspases, or even mitochondrial pathways that release cytochrome c, to induce target cell death. It is interesting that the 100K protein, which has important functions in the adenovirus life cycle, inhibits granzyme B by acting not as a serpin, but as a complex between a large amount of the inhibitor and a small amount of protease, which results in a slow rate of cleavage of the substrate. 100K protein is produced in adenovirus-infected cells at levels likely to be vastly in excess of the amounts of granzyme B that enter target cells during CTL-induced killing. The presence of this inhibitor probably allows adenovirusinfected cells to live several extra hours after CTL-mediated attack, which may buy the virus time to replicate. With additional time, granzyme A (and/or other mechanisms) still cause apoptosis of the target cell by using other biochemical pathways and substrates, which explains why CD8+ T cells ultimately have no difficulty in eliminating virtually all adenovirus infections. Nonetheless, this is an important demonstration of a virus that specifically targets granzyme B, implicating granzyme B as a clear and immediate threat to the adenovirus-infected cell.

THE DEVELOPMENT OF THERAPEUTIC INHIBITORS OF THE GRANULE EXOCYTOSIS PATHWAY The granule exocytosis pathway contains an enormous number of potential steps where inhibitors could be targeted. Granule formation, granule secretion, perforin insertion into the target cell membrane, the entry of granzymes, the intracellular trafficking of granzymes, or the granzyme molecules themselves could all be viewed as targets for inhibitors. The two clinical circumstances where inhibitors of this pathway could be utilized would be in GvHD and in autoimmune diseases like rheumatoid arthritis, where the granule exocytosis pathway has been implicated as a contributor to disease pathogenesis. Mouse models suggest that it would probably be very unwise to inhibit the granule exocytosis pathway when allogeneic transplants are being performed to provide a potent GvL effect. Under these circumstances, any benefit that could be realized by the reduction of GvHD could conceivably be lost by a higher relapse rate in the patients. However, the abrogation of GvHD for the treatment of nonmalignant disease would seem a more reasonable setting to test inhibitors of this pathway, if they could be developed. One newly identified risk of long-term inhibition of the granule exocytosis pathway, however, may be the development of hematophagocytic syndromes. Regardless, if inhibitors of this pathway were to be developed, it would seem prudent to target a position in the pathway where multiple granzymes could be

5 Feb 2002

12:52

AR

AR152-12.tex

AR152-12.SGM

LaTeX2e(2001/05/10)

P1: GJC

Annu. Rev. Immunol. 2002.20:323-370. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

LYMPHOCYTE-MEDIATED CYTOTOXICITY

353

inhibited. The loss-of-function models revealed that granzyme B cluster–deficiency resulted in some attenuation of CD8+-mediated GvHD (232), while granzyme A deficiency resulted in none (195). Combined deficiency was similar to that of perforin deficiency, and therefore, both granzymes probably would need to be inhibited in order for significant clinical benefit to occur (195). Specifically, inhibiting either granzyme A or B with a small molecule is still an extraordinarily difficult task (204); highly efficient inhibitors of these enzymes do not yet exist. This limits the possibilities for therapeutic intervention to altering the granule environment, inhibiting granule secretion, inhibiting perforin function (either directly, or via inhibition of perforin insertion into target cell membranes), inhibition of granzyme entry or trafficking, or pan-granzyme inhibition. Because the granzymes have different active sites and specificities, a limited number of options exist for broadly inhibiting granzyme function. One option is inhibition of DPPI, which is known to be required for the processing and activation of granzyme A and B (117). Clearly, a good DPPI inhibitor would also reduce the activity of mast cell chymases and several neutrophil azurophil granule proteinases (i.e., neutrophil elastase, cathepsin G, and proteinase 3). Remarkably, however, loss of function of this enzyme results in a relatively mild phenotype in human patients, and inhibition of this enzyme could therefore be tolerated. One small molecule inhibitor of DPPI has been developed, but it has not yet been tested in vivo (260). Finally, inhibition of the CI-MPR, suggested by Bleakley and colleagues, remains another possibility for exploration. This receptor can internalize granzyme B but is also expected to mediate internalization of granzyme A (and possibly the other granzymes as well). The CI-MPR receptor is widely expressed. It is not clear what would be required to develop an effective inhibitor of this enzyme, since the receptor is efficiently recycled after releasing ligand upon internalization. Although mice require this receptor during development, it is not required for adult life, and systemic inhibition could therefore presumably be tolerated. As noted above, however, the efficiency of CI-MPR blockade for granzyme inhibition will ultimately depend on whether this is a minor or a major pathway for granzyme uptake in vivo. The granule exocytosis pathway, first proposed by Henkart in this Annual Review in 1985 (97), has come a long way in 16 years. This pathway is extraordinarily intriguing and complex, and it has yielded tremendous biological information important for a variety of systems. The basic science that has led to the explosion of new facts about this pathway is still robust, and the translational potential of these discoveries will, we hope, be realized soon.

ACKNOWLEDGMENTS The authors dedicate this review to the memory of Arnold Greenberg. His many seminal contributions to this field are evident throughout this review. The authors thank Nancy Reidelberger for expert editorial assistance. This work was supported

5 Feb 2002

12:52

354

AR

AR152-12.tex

RUSSELL

¥

AR152-12.SGM

LaTeX2e(2001/05/10)

P1: GJC

LEY

by NIH AI45861 (JHR), National Multiple Sclerosis Society Grant RG2835 (JHR), NIH DK49786 (TJL), and the Alan and Edith L. Wolff Professorship (TJL). Visit the Annual Reviews home page at www.annualreviews.org

Annu. Rev. Immunol. 2002.20:323-370. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

LITERATURE CITED 1. Kagi D, Ledermann B, B¨urki K, Zinkernagel RM, Hengartner H. 1996. Molecular mechanisms of lymphocyte-mediated cytotoxicity and their role in immunological protection and pathogenesis in vivo. Annu. Rev. Immunol. 14:207–32 2. Kagi D, Vignaux F, Ledermann B, B¨urki K, Depraetere V, Nagata S, Hengartner H, Golstein P. 1994. Fas and perforin pathways as major mechanisms of T cellmediated cytotoxicity. Science 265:528– 30 3. Watanabe-Fukunaga R, Brannan CI, Copeland NG, Jenkins NA, Nagata S. 1992. Lymphoproliferation disorder in mice explained by defects in Fas antigen that mediates apoptosis. Nature 356:314–17 4. Takahashi T, Tanaka M, Brannan CI, Jenkins NA, Copeland NG, Suda T, Nagata S. 1994. Generalized lymphoproliferative disease in mice, caused by a point mutation in the Fas ligand. Cell 76:969– 76 5. Kagi D, Ledermann B, B¨urki K, Seiler P, Odermatt B, Olsen KJ, Podack ER, Zinkernagel RM, Hengartner H. 1994. Cytotoxicity mediated by T cells and natural killer cells is greatly impaired in perforindeficient mice. Nature 369:31–37 6. Kidd JG. 1950. Experimental necrobiosis—a venture in cytobiology. Proc. Inst. Med. Chicago 18:50–60 7. Russell JH, Masakowski VR, Dobos CB. 1980. Mechanisms of immune lysis. I. Physiological distinction between target cell death mediated by cytotoxic T lymphocytes and antibody plus complement. J. Immunol. 124:1100–5 8. Russell JH, Dobos CB. 1980. Mechanisms of immune lysis. II. CTL-induced

9.

10.

11.

12.

13.

14.

15.

nuclear disintegration of the target begins within minutes of cell contact. J. Immunol. 125:1256–61 Sellins KS, Cohen JJ. 1991. Cytotoxic T lymphocytes induce different types of DNA damage in target cells of different origins. J. Immunol. 147:795–803 Kuang AA, Diehl GE, Zhang J, Winoto A. 2000. FADD is required for DR4- and DR5-mediated apoptosis: lack of trailinduced apoptosis in FADD-deficient mouse embryonic fibroblasts. J. Biol. Chem. 275:25065–68 Bodmer JL, Holler N, Reynard S, Vinciguerra P, Schneider P, Juo P, Blenis J, Tschopp J. 2000. TRAIL receptor-2 signals apoptosis through FADD and caspase-8. Nat. Cell Biol. 2:241–43 Sprick MR, Weigand MA, Rieser E, Rauch CT, Juo P, Blenis J, Krammer PH, Walczak H. 2000. FADD/MORT1 and caspase-8 are recruited to TRAIL receptors 1 and 2 and are essential for apoptosis mediated by TRAIL receptor 2. Immunity 12:599–609 Kischkel FC, Lawrence DA, Chuntharapai A, Schow P, Kim KJ, Ashkenazi A. 2000. Apo2L/TRAIL-dependent recruitment of endogenous FADD and caspase8 to death receptors 4 and 5. Immunity 12:611–20 Doherty PC, Christensen JP. 2000. Accessing complexity: the dynamics of virusspecific T cell responses. Annu. Rev. Immunol. 18:561–92 Abrams SI, Brahmi Z. 1986. The functional loss of human natural killer cell activity induced by K562 is reversible via an interleukin-2-dependent mechanism. Cell. Immunol. 101:558–70

5 Feb 2002

12:52

AR

AR152-12.tex

AR152-12.SGM

LaTeX2e(2001/05/10)

P1: GJC

Annu. Rev. Immunol. 2002.20:323-370. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

LYMPHOCYTE-MEDIATED CYTOTOXICITY 16. Kupfer A, Dennert G, Singer SJ. 1985. The reorientation of the Golgi apparatus and the microtubule-organizing center in the cytotoxic effector cell is a prerequisite in the lysis of bound target cells. J. Mol. Cell. Immunol. 2:37–49 17. Podack ER, Young JD, Cohn ZA. 1985. Isolation and biochemical and functional characterization of perforin 1 from cytolytic T-cell granules. Proc. Natl. Acad. Sci. USA 82:8629–33 18. Zheng LX, Trageser CL, Willerford DM, Lenardo MJ. 1998. T cell growth cytokines cause the superinduction of molecules mediating antigen-induced T lymphocyte death. J. Immunol. 160:763–69 19. Ju ST, Cui H, Panka DJ, Ettinger R, Marshak-Rothstein A. 1994. Participation of target Fas protein in apoptosis pathway induced by CD4+ Th1 and CD8+ cytotoxic T cells. Proc. Natl. Acad. Sci. USA 91:4185–89 20. Nguyen T, Wang R, Russell JH. 2000. IL-12 enhances IL-2 function by inducing CD25 expression through a p38 mitogenactivated protein kinase pathway. Eur. J. Immunol. 30:1445–52 21. Wang R, Rogers AM, Ratliff TL, Russell JH. 1996. CD95-dependent bystander lysis caused by CD4+ T helper 1 effectors. J. Immunol. 157:2961–68 22. Williams NS, Engelhard VH. 1996. Identification of a population of CD4+ CTL that utilizes a perforin- rather than a Fas ligand-dependent cytotoxic mechanism. J. Immunol. 156:153–59 23. Yasukawa M, Ohminami H, Yakushijin Y, Arai J, Hasegawa A, Ishida Y, Fujita S. 1999. Fas-independent cytotoxicity mediated by human CD4+ CTL directed against herpes simplex virus-infected cells. J. Immunol. 162:6100–6 24. Schulz M, Schuurman HJ, Joergensen J, Steiner C, Meerloo T, Kagi D, Hengartner H, Zinkernagel RM, Schreier MH, Burki K. 1995. Acute rejection of vascular heart allografts by perforin-deficient mice. Eur. J. Immunol. 25:474–80

355

25. Graubert TA, DiPersio JF, Russell JH, Ley TJ. 1997. Perforin/granzyme-dependent and -independent mechanisms are both important for the development of graft-versus-host disease after murine bone marrow transplantation. J. Clin. Invest. 100:904–11 26. Van den Broek MF, Kagi D, Zinkernagel RM, Hengartner H. 1995. Perforin dependence of natural killer cell-mediated tumor control in vivo. Eur. J. Immunol. 25:3514–16 27. Meier P, Finch A, Evan G. 2000. Apoptosis in development. Nature 407:796–801 28. Rathmell JC, Townsend SE, Xu JC, Flavell RA, Goodnow CC. 1996. Expansion or elimination of B cells in vivo: dual roles for CD40- and Fas (CD95)-ligands modulated by the B cell antigen receptor. Cell 87:319–29 29. Rathmell JC, Thompson CB. 1999. The central effectors of cell death in the immune system. Annu. Rev. Immunol. 17: 781–828 30. Zinkernagel RM, Althage A. 1977. Antiviral protection by virus-immune cytotoxic T cells: Infected target cells are lysed before infectious virus progeny is assembled. J. Exp. Med. 145:644–51 31. Topham DJ, Tripp RA, Doherty PC. 1997. CD8+ T cells clear influenza virus by perforin or fas-dependent processes. J. Immunol. 159:5197–200 32. Parra B, Lin MT, Stohlman SA, Bergmann CC, Atkinson R, Hinton DR. 2000. Contributions of Fas-Fas ligand interactions to the pathogenesis of mouse hepatitis virus in the central nervous system. J. Virol. 74:2447–50 33. Parra B, Hinton DR, Marten NW, Bergmann CC, Lin MT, Yang CS, Stohlman SA. 1999. IFN-gamma is required for viral clearance from central nervous system oligodendroglia. J. Immunol. 162:1641– 47 34. Lin MT, Stohlman SA, Hinton DR. 1997. Mouse hepatitis virus is cleared from the central nervous systems of mice

5 Feb 2002

12:52

356

35.

Annu. Rev. Immunol. 2002.20:323-370. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

36.

37.

38.

39.

40.

41.

42.

43.

AR

AR152-12.tex

RUSSELL

¥

AR152-12.SGM

LaTeX2e(2001/05/10)

P1: GJC

LEY

lacking perforin-mediated cytolysis. J. Virol. 71:383–91 Kagi D, Seiler P, Pavlovic J, Ledermann B, Burki K, Zinkernagel RM, Hengartner H. 1995. The roles of perforin- and Fas-dependent cytotoxicity in protection against cytopathic and noncytopathic viruses. Eur. J. Immunol. 25:3256–62 Mullbacher A, Hla RT, Museteanu C, Simon MM. 1999. Perforin is essential for control of ectromelia virus but not related poxviruses in mice. J. Virol. 73:1665–67 Ghiasi H, Cai S, Perng G, Nesburn AB, Wechsler SL. 1999. Perforin pathway is essential for protection of mice against lethal ocular HSV-1 challenge but not corneal scarring. Virus Res. 65:97–101 Chang E, Galle L, Maggs D, Estes DM, Mitchell WJ. 2000. Pathogenesis of herpes simplex virus type 1-induced corneal inflammation in perforin-deficient mice. J. Virol. 74:11,832–40 Elkon KB, Liu CC, Gall JG, Trevejo J, Marino MW, Abrahamsen KA, Song X, Zhou JL, Old LJ, Crystal RG, FalckPedersen E. 1997. Tumor necrosis factor alpha plays a central role in immunemediated clearance of adenoviral vectors. Proc. Natl. Acad. Sci. USA 94:9814–19 Neumann B, Machleidt T, Lifka A, Pfeffer K, Vestweber D, Mak TW, Holzmann B, Kronke M. 1996. Crucial role of 55-kilodalton TNF receptor in TNFinduced adhesion molecule expression and leukocyte organ infiltration. J. Immunol. 156:1587–93 Kagi D, Ledermann B, Burki K, Hengartner H, Zinkernagel RM. 1994. CD8+ T cell-mediated protection against an intracellular bacterium by perforindependent cytotoxicity. Eur. J. Immunol. 24:3068–72 Jensen ER, Glass AA, Clark WR, Wing EJ, Miller JF, Gregory SH. 1998. Fas (CD95)-dependent cell-mediated immunity to Listeria monocytogenes. Infect. Immun. 66:4143–50 Nakane A, Minagawa T, Kato K. 1988.

44.

45.

46.

47.

48.

49.

50.

Endogenous tumor necrosis factor (cachectin) is essential to host resistance against Listeria monocytogenes infection. Infect. Immun. 56:2563–69 White DW, Badovinac VP, Kollias G, Harty JT. 2000. Cutting edge: antilisterial activity of CD8+ T cells derived from TNF-deficient and TNF/perforin doubledeficient mice. J. Immunol. 165:5– 9 Bruhn H, Leippe M. 1999. Comparative modeling of amoebapores and granulysin based on the NK-lysin structure-structural and functional implications. Biol. Chem. 380:1001–7 Pena SV, Hanson DA, Carr BA, Goralski TJ, Krensky AM. 1997. Processing, subcellular localization, and function of 519 (granulysin), a human late T cell activation molecule with homology to small, lytic, granule proteins. J. Immunol. 158:2680–88 Gamen S, Hanson DA, Kaspar A, Naval J, Krensky AM, Anel A. 1998. Granulysin-induced apoptosis. I. Involvement of at least two distinct pathways. J. Immunol. 161:1758–64 Stenger S, Hanson DA, Teitelbaum R, Dewan P, Niazi KR, Froelich CJ, Ganz T, Thoma-Uszynski S, Melian A, Bogdan C, Porcelli SA, Bloom BR, Krensky M, Modlin RL. 1998. An antimicrobial activity of cytolytic T cells mediated by granulysin. Science 282:121–25 Ochoa MT, Stenger S, Sieling PA, Thoma-Uszynski S, Sabet S, Cho S, Krensky AM, Rollinghoff M, Nunes Sarno E, Burdick AE, Rea TH, Modlin RL. 2001. T-cell release of granulysin contributes to host defense in leprosy. Nat. Med. 7:174– 79 Ernst WA, Thoma-Uszynski S, Teitelbaum R, Ko C, Hanson DA, Clayberger C, Krensky AM, Leippe M, Bloom BR, Ganz T, Modlin RL. 2000. Granulysin, a T cell product, kills bacteria by altering membrane permeability. J. Immunol. 165:7102–8

5 Feb 2002

12:52

AR

AR152-12.tex

AR152-12.SGM

LaTeX2e(2001/05/10)

P1: GJC

Annu. Rev. Immunol. 2002.20:323-370. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

LYMPHOCYTE-MEDIATED CYTOTOXICITY 51. Adachi M, Watanabe-Fukunaga R, Nagata S. 1993. Aberrant transcription caused by the insertion of an early transposable element in an intron of the Fas antigen gene of lpr mice. Proc. Natl. Acad. Sci. USA 90:1756–60 52. Rathmell JC, Cooke MP, Ho WY, Grein J, Townsend SE, Davis MM, Goodnow CC. 1995. CD95 (Fas)-dependent elimination of self-reactive B cells upon interaction with CD4+ T cells. Nature 376:181–84 53. Rothstein TL, Wang JK, Panka DJ, Foote LC, Wang Z, Stanger B, Cui H, Ju ST, Marshak-Rothstein A. 1995. Protection against Fas-dependent Th1-mediated apoptosis by antigen receptor engagement in B cells. Nature 374:163–65 54. Van Parijs L, Peterson DA, Abbas AK. 1998. The Fas/Fas ligand pathway and Bcl-2 regulate T cell responses to model self and foreign antigens. Immunity 8: 265–74 55. Russell JH, Wang R. 1993. Autoimmune gld mutation uncouples suicide and cytokine/proliferation pathways in activated, mature T cells. Eur. J. Immunol. 23:2379–82 56. Singer GG, Abbas AK. 1994. The fas antigen is involved in peripheral but not thymic deletion of T lymphocytes in T cell receptor transgenic mice. Immunity 1:365–71 57. Van Parijs L, Biuckians A, Abbas AK. 1998. Functional roles of Fas and Bcl-2regulated apoptosis of T lymphocytes. J. Immunol. 160:2065–71 58. de Saint Basile G, Fischer A. 2001. The role of cytotoxicity in lymphocyte homeostasis. Curr. Opin. Immunol. 13: In press 59. M´enasch´e G, Pastural E, Feldmann J, Certain S, Ersoy F, Dupuis S, Wulffraat N, Bianchi D, Fischer A, Le Deist F, de Saint Basile G. 2000. Mutations in RAB27A cause Griscelli syndrome associated with haemophagocytic syndrome. Nat. Genet. 25:173–76 60. Haddad EK, Wu X, Hammer JA III, Henkart PA. 2001. Defective granule ex-

61.

62.

63.

64.

65.

66.

67.

68.

69.

357

ocytosis in Rab27a-deficient lymphocytes from Ashen mice. J. Cell. Biol. 152:835– 42 Barbosa MD, Nguyen QA, Tchernev VT, Ashley JA, Detter JC, Blaydes SM, Brandt SJ, Chotai D, Hodgman C, Solari RC, Lovett M, Kingsmore SF. 1996. Identification of the homologous beige and Chediak-Higashi syndrome genes. Nature 382:262–65 Matloubian M, Suresh M, Glass A, Glavan M, Chow K, Whitmire JK, Walsh CM, Clark WR, Ahmed R. 1999. A role for perforin in downregulating T-cell responses during chronic viral infection. J. Virol. 73:2527–36 Sambhara S, Switzer I, Kurichh A, Miranda R, Urbanczyk L, James O, Underdown B, Klein M, Burt D. 1998. Enhanced antibody and cytokine responses to influenza viral antigens in perforindeficient mice. Cell. Immunol. 187:13– 18 Peng SL, Moslehi J, Robert ME, Craft J. 1998. Perforin protects against autoimmunity in lupus-prone mice. J. Immunol. 160:652–60 Kagi D, Odermatt B, Mak TW. 1999. Homeostatic regulation of CD8+ T cells by perforin. Eur. J. Immunol. 29:3262–72 Zhou T, Edwards CK 3rd, Yang P, Wang Z, Bluethmann H, Mountz JD. 1996. Greatly accelerated lymphadenopathy and autoimmune disease in lpr mice lacking tumor necrosis factor receptor I. J. Immunol. 156:2661–65 Zheng L, Fisher G, Miller RE, Peschon J, Lynch DH, Lenardo MJ. 1995. Induction of apoptosis in mature T cells by tumour necrosis factor. Nature 377:348–51 Chan FK, Lenardo MJ. 2000. A crucial role for p80 TNF-R2 in amplifying p60 TNF-R1 apoptosis signals in T lymphocytes. Eur. J. Immunol. 30:652–60 Teh HS, Seebaran A, Teh SJ. 2000. TNF receptor 2-deficient CD8 T cells are resistant to Fas/Fas ligand-induced cell death. J. Immunol. 165:4814–21

5 Feb 2002

12:52

Annu. Rev. Immunol. 2002.20:323-370. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

358

AR

AR152-12.tex

RUSSELL

¥

AR152-12.SGM

LaTeX2e(2001/05/10)

P1: GJC

LEY

70. Nguyen LT, McKall-Faienza K, Zakarian A, Speiser DE, Mak TW, Ohashi PS. 2000. TNF receptor 1 (TNFR1) and CD95 are not required for T cell deletion after virus infection but contribute to peptideinduced deletion under limited conditions. Eur. J. Immunol. 30:683–88 71. Reich A, Korner H, Sedgwick JD, Pircher H. 2000. Immune down-regulation and peripheral deletion of CD8 T cells does not require TNF receptor-ligand interactions nor CD95 (Fas, APO-1). Eur. J. Immunol. 30:678–82 72. Kreuwel HTC, Morgan DG, Krahl T, Ko A, Sarvetnick N, Sherman LA. 1999. Comparing the relative role of perforin/ granzyme versus fas/fas ligand cytotoxic pathways in CD8+ T cell-mediated insulin-dependent diabetes mellitus. J. Immunol. 163:4335–41 73. Kagi D, Odermatt B, Ohashi PS, Zinkernagel RM, Hengartner H. 1996. Development of insulitis without diabetes in transgenic mice lacking perforin-dependent cytotoxicity. J. Exp. Med. 183:2143–52 74. Chervonsky AV, Wang Y, Wong FS, Visintin I, Flavell RA, Janeway CA, Matis LA. 1997. The role of Fas in autoimmune diabetes. Cell 89:17–24 75. Herrera PL, Harlan DM, Vassalli P. 2000. A mouse CD8 T cell-mediated acute autoimmune diabetes independent of the perforin and Fas cytotoxic pathways: possible role of membrane TNF. Proc. Natl. Acad. Sci. USA 97:279–84 76. Kagi D, Odermatt B, Seiler P, Zinkernagel RM, Mak TW, Hengartner H. 1997. Reduced incidence and delayed onset of diabetes in perforin-deficient nonobese diabetic mice. J. Exp. Med. 186:989–97 77. Su X, Hu Q, Kristan JM, Costa C, Shen Y, Gero D, Matis LA, Wang Y. 2000. Significant role for Fas in the pathogenesis of autoimmune diabetes. J. Immunol. 164:2523–32 78. Pakala SV, Chivetta M, Kelly CB, Katz JD. 1999. In autoimmune diabetes the transition from benign to pernicious in-

79.

80.

81.

82.

83.

84.

85.

86.

sulitis requires an islet cell response to tumor necrosis factor alpha. J. Exp. Med. 189:1053–62 Gebhard JR, Perry CM, Harkins S, Lane T, Mena I, Asensio VC, Campbell IL, Whitton JL. 1998. Coxsackievirus B3induced myocarditis: Perforin exacerbates disease, but plays no detectable role in virus clearance. Am. J. Pathol. 153:417–28 Hogquist KA, Nett MA, Unanue ER, Chaplin DD. 1991. Interleukin 1 is processed and released during apoptosis. Proc. Natl. Acad. Sci. USA 88:8485–89 Malipiero U, Frei K, Spanaus K-S, Agresti C, Lassmann H, Hahne M, Tschopp J, Eugster H-P, Fontana A. 1997. Myelin oligodendrocyte glycoprotein-induced autoimmune encephalomyelitis is chronic/relapsing in perforin knockout mice, but monophasic in Fas- and Fas liganddeficient lpr and gld mice. Eur. J. Immunol. 27:3151–60 Sabelko KA, Kelly KA, Nahm MH, Cross AH, Russell JH. 1997. Fas and Fas ligand enhance the pathogenesis of experimental allergic encephalomyelitis, but are not essential for immune privilege in the central nervous system. J. Immunol. 159:3096–99 Waldner H, Sobel RA, Howard E, Kuchroo VK. 1997. Fas- and FasL-deficient mice are resistant to induction of autoimmune encephalomyelitis. J. Immunol. 159:3100–3 Okuda Y, Bernard CC, Fujimura H, Yanagihara T, Sakoda S. 1998. Fas has a crucial role in the progression of experimental autoimmune encephalomyelitis. Mol. Immunol. 35:317–26 Elliott EA, McFarland HI, Nye SH, Cofiell R, Wilson TM, Wilkins JA, Squinto SP, Matis LA, Mueller JP. 1996. Treatment of experimental encephalomyelitis with a novel chimeric fusion protein of myelin basic protein and proteolipid protein. J. Clin. Invest. 98:1602–12 Suvannavejh GC, Dal Canto MC, Matis

5 Feb 2002

12:52

AR

AR152-12.tex

AR152-12.SGM

LaTeX2e(2001/05/10)

P1: GJC

LYMPHOCYTE-MEDIATED CYTOTOXICITY

Annu. Rev. Immunol. 2002.20:323-370. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

87.

88.

89.

90.

91.

92.

93.

94.

LA, Miller SD. 2000. Fas-mediated apoptosis in clinical remissions of relapsing experimental autoimmune encephalomyelitis. J. Clin. Invest. 105:223–31 Sabelko-Downes KA, Gimenez MT, Suvannavejh GC, Miller SD, Russell JH. 2000. Genetic control of pathogenic mechanisms in autoimmune demyelinating disease. J. Neuroimmunol. 110:168– 76 Liu TS, Hilliard B, Samoilova EB, Chen Y. 2000. Differential roles of Fas ligand in spontaneous and actively induced autoimmune encephalomyelitis. Clin. Immunol. 95:203–11 Sabelko-Downes KA, Cross AH, Russell JH. 1999. Dual role for Fas ligand in the initiation of and recovery from experimental allergic encephalomyelitis. J. Exp. Med. 189:1195–205 Griffiths GM, Alpert S, Lambert E, McGuire J, Weissman IL. 1992. Perforin and granzyme A expression identifying cytolytic lymphocytes in rheumatoid arthritis. Proc. Natl. Acad. Sci. USA 89:549–53 Seko Y, Minota S, Kawasaki A, Shinkai Y, Maeda K, Yagita H, Okumura K, Sato O, Takagi A, Tada Y, Yazaki Y. 1994. Perforin-secreting killer cell infiltration and expression of a 65-kD heatshock protein in aortic tissue of patients with Takayasu’s arteritis. J. Clin. Invest. 93:750–58 Giordano C, Stassi G, Demaria R, Todaro M, Richiusa P, Papoff G, Ruberti G, Bagnasco M, Testi R, Galluzzo A. 1997. Potential involvement of Fas and its ligand in the pathogenesis of Hashimoto’s thyroiditis. Science 275:960–63 Trautmann A, Akdis M, Kleemann D, Altznauer F, Simon HU, Graeve T, Noll M, Brocker EB, Blaser K, Akdis CA. 2000. T cell-mediated Fas-induced keratinocyte apoptosis plays a key pathogenetic role in eczematous dermatitis. J. Clin. Invest. 106:25–35 Dowling P, Shang G, Raval S, Menonna J, Cook S, Husar W. 1996. Involvement

95.

96.

97.

98.

99.

100.

101.

102.

103.

104.

359

of the CD95 (APO-1/Fas) receptor/ligand system in multiple sclerosis brain. J. Exp. Med. 184:1513–18 D’Souza SD, Bonetti B, Balasingam V, Cashman NR, Barker PA, Troutt AB, Raine CS, Antel JP. 1996. Multiple sclerosis: Fas signaling in oligodendrocyte cell death. J. Exp. Med. 184:2361–70 Smith D, Sieg S, Kaplan D. 1998. Technical note: aberrant detection of cell surface Fas ligand with anti-peptide antibodies. J. Immunol. 160:4159–60 Henkart PA. 1985. Mechanism of lymphocyte-mediated cytotoxicity. Annu. Rev. Immunol. 3:31–58 Shresta S, Pham CTN, Thomas DA, Graubert TA, Ley TJ. 1998. How do cytotoxic lymphocytes kill their target cells? Curr. Opin. Immunol. 10:581–87 Shiver JW, Henkart PA. 1991. A noncytotoxic mast cell tumor line exhibits potent IgE-dependent cytotoxicity after transfection with the cytolysin/perforin gene. Cell 64:1175–81 Duke RC, Persechini PM, Chang S, Liu C-C, Cohen J, Young JD-E. 1989. Purified perforin induces target cell lysis but not DNA fragmentation. J. Exp. Med. 170:1451–56 Shi L, Kam C-M, Powers JC, Aebersold R, Greenberg AH. 1992. Purification of three cytotoxic lymphocyte granule serine proteases that induce apoptosis through distinct substrate and target cell interactions. J. Exp. Med. 176:1521–29 Shi L, Kraut RP, Aebersold R, Greenberg AH. 1992. A natural killer cell granule protein that induces DNA fragmentation and apoptosis. J. Exp. Med. 175:553–66 Shiver JW, Su L, Henkart PA. 1992. Cytotoxicity with target DNA breakdown by rat basophilic leukemia cells expressing both cytolysin and granzyme A. Cell 71:315–22 Nakajima H, Park HL, Henkart PA. 1995. Synergistic roles of granzymes A and B in mediating target cell death by rat basophilic leukemia mast cell tumors

5 Feb 2002

12:52

360

105.

Annu. Rev. Immunol. 2002.20:323-370. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

106.

107.

108.

109.

110.

111.

112.

113.

AR

AR152-12.tex

RUSSELL

¥

AR152-12.SGM

LaTeX2e(2001/05/10)

P1: GJC

LEY

also expressing cytolysin/perforin. J. Exp. Med. 181:1037–46 Podack ER. 1999. How to induce involuntary suicide: the need for dipeptidyl peptidase I. Proc. Natl. Acad. Sci. USA 96:8312–14 Heusel JW, Wesselschmidt R, Shresta S, Russell JH, Ley TJ. 1994. Cytotoxic lymphocytes require granzyme B for the rapid induction of DNA fragmentation and apoptosis in allogeneic target cells. Cell 76:977–87 Shresta S, MacIvor DM, Heusel JW, Russell JH, Ley TJ. 1995. Natural killer and lymphokine-activated-killer cells require granzyme B for the rapid induction of DNA fragmentation and apoptosis in susceptible target cells. Proc. Natl. Acad. Sci. USA 92:5679–83 Odake S, Kam C-M, Narasimhan L, Poe M, Blake JT, Krahenbuhl O, Tschopp J, Powers JC. 1991. Human and murine cytotoxic T lymphocyte serine proteases: subsite mapping with peptide thioester substrates and inhibition of enzyme activity and cytolysis by isocoumarins. Biochemistry 30:2217–27 Darmon AJ, Nicholson DW, Bleackley RC. 1995. Activation of the apoptotic protease CPP32 by cytotoxic T-cell-derived granzyme B. Nature 377:446–48 Quan LT, Tewari M, O’Rourke K, Dixit V, Snipas SJ, Poirier GG, Ray C, Pickup DJ, Salvesen GS. 1996. Proteolytic activation of the cell death protease Yama/ CPP32 by granzyme B. Proc. Natl. Acad. Sci. USA 93:1972–76 Darmon AJ, Ley TJ, Nicholson DW, Bleackley RC. 1996. Cleavage of CPP32 by granzyme B represents a critical role for granzyme B in the induction of target cell DNA fragmentation. J. Biol. Chem. 271:21709–12 Stepp SE, Mathew PA, Bennett M, de Saint Basile G, Kumar V. 2000. Perforin: more than just an effector molecule. Trends Immunol. Today 21:254–56 Caputo A, Garner RS, Winkler U, Hudig

114.

115.

116.

117.

118.

119.

120.

121.

122.

D, Bleackley RC. 1993. Activation of recombinant murine cytotoxic cell proteinase-1 requires deletion of an aminoterminal dipeptide. J. Biol. Chem. 268: 17672–75 McGuire MJ, Lipsky PE, Thiele DL. 1993. Generation of active myeloid and lymphoid granule serine proteases requires processing by the granule thiol protease dipeptidyl peptidase I. J. Biol. Chem. 268:2458–67 Smyth MJ, McGuire MJ, Thia KYT. 1995. Expression of recombinant human granzyme B. J. Immunol. 154:6299–305 Pham CTN, Thomas DA, Mercer JD, Ley TJ. 1998. Production of fully active recombinant murine granzyme B in yeast. J. Biol. Chem. 273:1629–33 Pham CTN, Ley TJ. 1999. Dipeptidyl peptidase I is required for the processing and activation of granzymes A and B in vivo. Proc. Natl. Acad. Sci. USA 96:8627–32 Griffiths GM, Isaaz S. 1993. Granzymes A and B are targeted to the lytic granules of lymphocytes by the mannose-6phosphate receptor. J. Cell Biol. 120:885– 96 Griffiths GM. 1997. Protein sorting and secretion during CTL killing. Semin. Immunol. 9:109–15 Uellner R, Zvelebil MJ, Hopkins J, Jones J, MacDougall LK, Morgan BP, Podack E, Waterfield MD, Griffiths GM. 1997. Perforin is activated by a proteolytic cleavage during biosynthesis which reveals a phospholipid-binding C2 domain. EMBO J. 16:7287–96 Kataoka T, Togashi K, Takayama H, Takaku K, Nagai K. 1997. Inactivation and proteolytic degradation of perforin within lytic granules upon neutralization of acidic pH. Immunology 91:493–500 Taylor MA, Bennett M, Kumar V, Schatzle JD. 2000. Functional defects of NK cells treated with chloroquine mimic the lytic defects observed in perforindeficient mice. J. Immunol. 165:5048–53

5 Feb 2002

12:52

AR

AR152-12.tex

AR152-12.SGM

LaTeX2e(2001/05/10)

P1: GJC

Annu. Rev. Immunol. 2002.20:323-370. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

LYMPHOCYTE-MEDIATED CYTOTOXICITY 123. Galvin JP, Spaeny-Dekking LHA, Wang B, Seth P, Hack CE, Froelich CJ. 1999. Apoptosis induced by granzyme B-glycosaminoglycan complexes: implications for granule-mediated apoptosis in vivo. J. Immunol. 162:5345–50 124. Thomas DA, Du C, Xu M, Wang X, Ley TJ. 2000. DFF45/ICAD can be directly processed by granzyme B during the induction of apoptosis. Immunity 12:621–32 125. Fraser SA, Karimi R, Michalak M, Hudig D. 2000. Perforin lytic activity is controlled by calreticulin. J. Immunol. 164:4150–55 126. Bossi G, Griffiths GM. 1999. Degranulation plays an essential part in regulating cell surface expression of Fas ligand in T cells and natural killer cells. Nat. Med. 5:90–96 127. Krensky AM. 2000. Granulysin: a novel antimicrobial peptide of cytolytic T lymphocytes and natural killer cells. Biochem. Pharmacol. 59:317–20 128. Kaspar AA, Okada S, Kumar J, Poulain FR, Drouvalakis KA, Kelekar A, Hanson DA, Kluck RM, Hitoshi Y, Johnson DE, Froelich CJ, Thompson CB, Newmeyer DD, Anel A, Clayberger C, Krensky AM. 2001. A distinct pathway of cell-mediated apoptosis initiated by granulysin. J. Immunol. 167:350–56 129. Lowin B, Hahne M, Mattmann C, Tschopp J. 1994. Cytolytic T-cell cytotoxicity is mediated through perforin and fas lytic pathways. Nature 370:650–52 130. Shresta S, Russell JH, Ley TJ. 1997. Mechanisms responsible for granzyme Bindependent cytotoxicity. Blood 89:4085– 91 131. Barbosa MDFS, Barrat FJ, Tchernev VT, Nguyen QA, Mishra VS, Colman SD, Pastural E, Dufourcq-Lagelouse R, Fischer A, Holcombe RF, Wallace MR, Brandt SJ, de Saint Basile G, Kingsmore SF. 1997. Identification of mutation in two major mRNA isoforms of the ChediakHigashi syndrome. Hum. Mol. Genet. 6:1091–98

361

132. Nagle DL, Karim MA, Woolf EA, Holmgren L, Bork P, Misumi DJ, McGrail SH, Dussault BJ Jr, Perou CM, Boissy RE, Duyk GM, Spritz RA, Moore KH. 1996. Identification and mutation analysis of the complete gene for Chediak-Higashi syndrome. Nat. Genet. 14:307–11 133. Spritz RA. 1998. Genetic defects in Chediak-Higashi syndrome and the beige mouse. J. Clin. Immunol. 18:97–105 134. Wei S, Gamero AM, Liu JH, Daulton AA, Valkov NI, Trapani JA, Larner AC, Weber MJ, Djeu JY. 1998. Control of lytic function by mitogen-activated protein kinase/extracellular regulatory kinase 2 (ERK2) in a human natural killer cell line: identification of perforin and granzyme B mobilization by functional ERK2. J. Exp. Med. 187:1753–65 135. Wei S, Gilvary DL, Corliss BC, Sebti S, Sun J, Straus DB, leibson PJ, Trapani JA, Hamilton AD, Weber MJ, Djeu JY. 2000. Direct tumor lysis by NK cells uses a Ras-independent mitogen-activated protein kinase signal pathway. J. Immunol. 165:3811–19 136. Jiang K, Zhong B, Gilvary DL, Corliss BC, Hong-Geller E, Wei S, Djeu JY. 2000. Pivotal role of phosphoinositide3 kinase in regulation of cytotoxicity in natural killer cells. Nat. Immunol. 1:419– 25 137. Stinchcome JC, Barral DC, Mules EH, Booth S, Hume AN, Manchesky LM, Seabra MC, Griffiths GM. 2001. Rab27a is required for regulated secretion in cytotoxic T lymphocytes. J. Cell Biol. 152:825–34 138. Browne KA, Blink E, Sutton VR, Froelich CJ, Jans DA, Trapani JA. 1999. Cytosolic delivery of granzyme B by bacterial toxins: evidence that endosomal disruption, in addition to transmembrane pore formation, is an important function of perforin. Mol. Cell. Biol. 19:8604–15 139. Tschopp J, Schafer S, Masson D, Peitsch MC, Heusser C. 1989. Phosphorylcholine acts as a Ca++ dependent receptor

5 Feb 2002

12:52

362

140.

Annu. Rev. Immunol. 2002.20:323-370. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

141.

142.

143.

144.

145.

146.

147.

AR

AR152-12.tex

RUSSELL

¥

AR152-12.SGM

LaTeX2e(2001/05/10)

P1: GJC

LEY

molecule for lymphocyte perforin. Nature 337:272–74 M¨uller C, Tschopp J. 1994. Resistance of CTL to perforin-mediated lysis. J. Immunol. 153:2470–78 Rochel N, Cowan JA. 1997. Dependence of the lytic activity of the N-terminal domain of human perforin on membrane lipid composition. Implications for Tcell self-preservation. Eur. J. Biochem. 249:223–31 Berthou C, Bourge J-F, Zhang Y, Souli´e A, Geromin D, Denizot Y, Sigaux F, Sasportes M. 2000. Interferon-γ -induced membrane PAF-receptor expression confers tumor cell susceptibility to NK perforin-dependent lysis. Blood 95:2329– 36 Lehmann C, Zeis M, Schmitz N, Uharek L. 2000. Impaired binding of perforin on the surface of tumor cells is a cause of target cell resistance against cytotoxic effector cells. Blood 96:594–600 Podack ER, Lowrey DM, Lichtenheld MG, Hameed A. 1988. Function of granule perforin and esterases in T cellmediated reactions. Components required for delivery of molecules to target cells. Ann. NY Acad. Sci. 532:292–302 Froelich CJ, Orth K, Turbov J, Seth P, Gottlieb R, Babior B, Shah GM, Bleackley RC, Dixit VM, Hanna W. 1996. New paradigm for lymphocyte granulemediated cytotoxicity. Target cells bind and internalize granzyme B, but an endosomolytic agent is necessary for cytosolic delivery and subsequent apoptosis. J. Biol. Chem. 271:29073–79 Shi L, Mai S, Israels S, Browne K, Trapani JA, Greenberg AH. 1997. Granzyme B (GraB) autonomously crosses the cell membrane and perforin initiates apoptosis and GraB nuclear localization. J. Exp. Med. 185:855–66 Pinkoski MJ, Hobman M, Heibein JA, Tomaselli K, Li F, Seth P, Froelich CJ, Bleackley RC. 1998. Entry and trafficking of granzyme B in target cells during

148.

149.

150.

151.

152.

153.

154.

155.

granzyme B-perforin-mediated apoptosis. Blood 92:1044–54 Jans DA, Jans P, Briggs LJ, Sutton V, Trapani JA. 1996. Nuclear transport of granzyme B (Fragmentin-2). J. Biol. Chem. 271:30781–89 Trapani JA, Browne KA, Smyth MJ, Jans DA. 1996. Localization of granzyme B in the nucleus. J. Biol. Chem. 271:4127–33 Trapani JA, Jans P, Smyth MJ, Froelich CJ, Williams EA, Sutton VR, Jans DA. 1998. Perforin-dependent nuclear entry of granzyme B precedes apoptosis, and is not a consequence of nuclear membrane dysfunction. Cell Death Differ. 5:488– 96 Jans DA, Briggs LJ, Jans P, Froelich CJ, Parasivam G, Kumar S, Sutton VR, Trapani JA. 1998. Nuclear targeting of the serine protease granzyme A (fragmentin1). J. Cell Sci. 111:2645–54 Motyka B, Korbutt G, Pinkoski MJ, Heibein JA, Caputo A, Hobman M, Barry M, Shostak I, Sawchuk T, Holmes CFB, Gauldie J, Bleackley RC. 2000. Mannose 6-phosphate/insulin-like growth factor II receptor is a death receptor for granzyme B during cytotoxic T cell-induced apoptosis. Cell 103:491–500 Xia Z, Kam C-M, Huang C, Powers JC, Mandle RJ, Stevens RL, Lieberman J. 1998. Expression and purification of enzymatically active recombinant granzyme B in a baculovirus system. Biochem. Biophys. Res. Commun. 243:384–89 Thornberry NA, Rano TA, Peterson EP, Rasper DM, Timkey T, Garcia-Calvo M, Houtzager VM, Nordstrom PA, Roy S, Vaillancourt JP, Chapman KT, Nicholson DW. 1997. A combinatorial approach defines specificities of members of the caspase family and granzyme B. J. Biol. Chem. 272:17907–11 Harris JL, Peterson EP, Hudig D, Thornberry NA, Craik CS. 1998. Definition and redesign of the extended substrate specificity of granzyme B. J. Biol. Chem. 273:27,364–73

5 Feb 2002

12:52

AR

AR152-12.tex

AR152-12.SGM

LaTeX2e(2001/05/10)

P1: GJC

Annu. Rev. Immunol. 2002.20:323-370. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

LYMPHOCYTE-MEDIATED CYTOTOXICITY 156. Waugh SM, Harris JL, Fletterick R, Craik CS. 2000. The structure of the proapoptotic protease granzyme B reveals the molecular determinants of its specificity. Nat. Struct. Biol. 7:762–65 157. Estebanez-Perpina E, Fuentes-Prior P, Belorgey D, Braun M, Kiefersauer R, Maskos K, Huber R, Rubin H, Bode W. 2000. Crystal structure of the caspase activator human granzyme B, a proteinase highly specific for an Asp-P1 residue. J. Biol. Chem. 381:1203–14 158. Rotonda J, Garcia-Calvo M, Bull HG, Geissler WM, McKeever BM, Willoughby CA, Thornberry NA, Becker JW. 2001. The three-dimensional structure of human granzyme B compared to caspase3, key mediators of cell death with cleavage specificity for aspartic acid in P1. Chem. Biol. 8:357–68 159. Duan H, Orth K, Chinnaiyan AM, Poirier GG, Froelich CJ, He W-W, Dixit VM. 1996. ICE-LAP6, a novel member of the ICE/Ced-3 gene family, is activated by the cytotoxic T cell protease granzyme B. J. Biol. Chem. 271:16,720–24 160. Martin SJ, Amarante-Mendes GP, Shi L, Chuang T-H, Casiano CA, O’Brien GA, Fitzgerald P, Tan EM, Bokoch GM, Greenberg AH, Green DR. 1996. The cytotoxic cell protease granzyme B initiates apoptosis in a cell-free system by proteolytic processing and activation of the ICE/CED-3 family protease, CPP32, via a novel two-step mechanism. EMBO J. 15:2407–16 161. Shi L, Chen G, MacDonald G, Bergeron L, Li H, Miura M, Rotello RJ, Miller DK, Li P, Seshadri T, Yuan J, Greenberg AH. 1996. Activation of an interleukin 1 converting enzyme-dependent apoptosis pathway by granzyme B. Proc. Natl. Acad. Sci. USA 93:11002–7 162. Talanian RV, Yang X, Turbov J, Seth P, Ghayur T, Casiano CA, Orth K, Froelich CJ. 1997. Granule-mediated killing: pathways for granzyme B-initiated apoptosis. J. Exp. Med. 186:1323–31

363

163. Yang X, Stennicke HR, Wang B, Green DR, J¨anicke RU, Srinivasan A, Seth P, Salvesen GS, Froelich CJ. 1998. Granzyme B mimics apical caspases. J. Biol. Chem. 273:34,278–83 164. Sarin A, Williams MS, Alexander-Miller MA, Berzofsky JA, Zacharchuk CM, Henkart PA. 1997. Target cell lysis by CTL granule exocytosis is independent of ICE/Ced-3 family proteases. Immunity 6:209–15 165. Sarin A, Haddad E, Henkart PA. 1998. Caspase dependence of target cell damage induced by cytotoxic lymphocytes. J. Immunol. 161:2810–16 166. Nakajima H, Golstein P, Henkart PA. 1995. The target cell nucleus is not required for cell-mediated granzyme- or Fas-based cytotoxicity. J. Exp. Med. 181: 1905–9 167. Barry M, Heibein JA, Pinkoski MJ, Lee S-F, Moyer RW, Green DR, Bleackley RC. 2000. Granzyme B short-circuits the need for caspase 8 activity during granulemediated cytotoxic T-lymphocyte killing by directly cleaving Bid. Mol. Cell. Biol. 20:3781–94 168. MacDonald G, Shi L, Vande Velde C, Lieberman J, Greenberg AH. 1999. Mitochondria-dependent and -independent regulation of granzyme B-induced apoptosis. J. Exp. Med. 189:131–43 169. Sutton VR, Davis JE, Cancilla M, Johnstone RW, Ruefli AA, Sedelies K, Browne KA, Trapani JA. 2000. Initiation of apoptosis by granzyme B requires direct cleavage of Bid, but not direct granzyme B-mediated caspase activation. J. Exp. Med. 192:1403–13 170. Liu X, Zou H, Slaughter C, Wang X. 1997. DFF, a heterodimeric protein that functions downstream of caspase-3 to trigger DNA fragmentation during apoptosis. Cell 89:175–84 171. Liu X, Li P, Widlak P, Zou H, Luo X, Garrard WT, Wang X. 1998. The 40kDa subunit of DNA fragmentation factor induces DNA fragmentation and

5 Feb 2002

12:52

364

172.

Annu. Rev. Immunol. 2002.20:323-370. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

173.

174.

175.

176.

177.

178.

179.

180.

AR

AR152-12.tex

RUSSELL

¥

AR152-12.SGM

LaTeX2e(2001/05/10)

P1: GJC

LEY

chromatin condensation during apoptosis. Proc. Natl. Acad. Sci. USA 95:8461–66 Enari M, Sakahira H, Yokoyama H, Okawa K, Iwamatsu A, Nagata S. 1998. A caspase-activated DNase that degrades DNA during apoptosis and its inhibitor ICAD. Nature 391:43–50 Sakahira H, Enari M, Nagata S. 1998. Cleavage of CAD inhibitor in CAD activation and DNA degradation during apoptosis. Nature 391:96–99 Wolf BB, Schuler M, Echeverri F, Green DR. 1999. Caspase-3 is the primary activator of apoptotic DNA fragmentation via DNA fragmentation factor-45/inhibitor of caspase-activated DNase inactivation. J. Biol. Chem. 274:30651–56 Zhang J, Liu X, Scherer DC, Van Kaer L, Wang X, Xu M. 1998. Resistance to DNA fragmentation and chromatin condensation in mice lacking the DNA fragmentation factor 45. Proc. Natl. Acad. Sci. USA 95:12,480–85 Zhang J, Wang X, Bove KE, Xu M. 1999. DNA fragmentation factor 45-deficient cells are more resistant to apoptosis and exhibit different dying morphology than wild-type control cells. J. Biol. Chem. 274:37,450–54 Sharif-Askari E, Alam A, Rh´eaume E, Beresford PJ, Scotto C, Sharma K, Lee D, DeWolf WE, Nuttall ME, Lieberman J, S´ekaly R-P. 2001. Direct cleavage of the human DNA fragmentation factor-45 by granzyme B induces caspase-activated DNase release and DNA fragmentation. EMBO J. 20:3101–13 Chiu VK, Walsh CM, Liu CC, Reed JC, Clark WR. 1995. Bcl-2 blocks degranulation but not fas-based cell-mediated cytotoxicity. J. Immunol. 154:2023–32 Sutton VR, Vaux DL, Trapani JA. 1997. Bcl-2 prevents apoptosis induced by perforin and granzyme B, but not that mediated by whole cytotoxic lymphocytes. J. Immunol. 158:5783–90 Renvoize C, Roger R, Moulian N, Bertoglio J, Breard J. 1997. Bcl-2 expression in

181.

182.

183.

184.

185.

186.

187.

target cells leads to functional inhibition of caspase-3 protease family in human NK and lymphokine-activated killer cell granule-mediated apoptosis. J. Immunol. 159:126–34 Jans DA, Sutton VR, Jans P, Froelich CJ, Trapani JA. 1999. Bcl-2 blocks perforininduced nuclear translocation of granzymes concomitant with protection against the nuclear events of apoptosis. J. Biol. Chem. 274:3953–61 Davis JE, Sutton VR, Smyth MJ, Trapani JA. 2000. Dependence of granzyme B-mediated cell death on a pathway regulated by Bcl-2 or its viral homolog, BHRF1. Cell Death Differ. 7:973– 83 Pinkoski MJ, Waterhouse NJ, Heibein JA, Wolf BB, Kuwana T, Goldstein JC, Newmeyer DD, Bleackley RC, Green DR. 2001. Granzyme B-mediated apoptosis proceeds predominantly through a Bcl-2-inhibitable mitochondrial pathway. J. Biol. Chem. 276:12060–67 Heibein JA, Goping IS, Barry M, Pinkoski MJ, Shore GC, Green DR, Bleackley RC. 2000. Granzyme B-mediated cytochrome c release is regulated by the Bcl-2 family members Bid and Bax. J. Exp. Med. 192:1391–1401 Alimonti JB, Shi L, Baijal PK, Greenberg AH. 2001. Granzyme B induces Bidmediated cytochrome c release and mitochondrial permeability transition. J. Biol. Chem. 276:6974–82 Thomas DA, Scorrano L, Putcha GV, Korsmeyer SJ, Ley TJ. 2001. Granzyme B can cause mitochondrial depolarization and cell death in the absence of Bid, Bax, Bak or cytochrome c release. Proc. Natl. Acad. Sci. USA. In press Froelich CJ, Hanna WL, Poirier GG, Duriez PJ, D’Amours D, Salvesen GS, Alnemri ES, Earnshaw WC, Shah GM. 1996. Granzyme B/perforin-mediated apoptosis of Jurkat cells results in cleavage of Poly(ADP-ribose polymerase to the 89-kDa apoptotic fragment and

5 Feb 2002

12:52

AR

AR152-12.tex

AR152-12.SGM

LaTeX2e(2001/05/10)

P1: GJC

LYMPHOCYTE-MEDIATED CYTOTOXICITY

188.

Annu. Rev. Immunol. 2002.20:323-370. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

189.

190.

191.

192.

193.

194.

195.

196.

less abundant 64-kDa fragment. Biochem. Biophys. Res. Commun. 227:658–65 Andrade F, Roy S, Nicholson D, Thornberry N, Rosen A, Casciola-Rosen L. 1998. Granzyme B directly and efficiently cleaves several downstream caspase substrates: implications for CTLinduced apoptosis. Immunity 8:451–60 Browne KA, Johnstone RW, Jans DA, Trapani JA. 2000. Filamin (280-kDa actin-binding protein) is a caspase substrate and is also cleaved directly by cytotoxic T lymphocyte protease granzyme B during apoptosis. J. Biol. Chem. 275: 39262–66 Froelich CJ, Zhang X, Turbov J, Hudig D, Winkler U, Hanna WL. 1993. Human granzyme B degrades aggrecan proteoglycan in matrix synthesized by chondrocytes. J. Immunol. 151:7161–71 Zhang D, Beresford PJ, Greenberg AH, Lieberman J. 2001. Granzymes A and B directly cleave lamins and disrupt the nuclear lamina during granule-mediated cytolysis. Proc. Natl. Acad. Sci. USA 98: 5746–51 Casciola-Rosen L, Andrade F, Ulanet D, Wong DB, Rosen A. 1999. Cleavage by granzyme B is strongly predictive of autoantigen status: implications for initiation of autoimmunity. J. Exp. Med. 190:815–26 Pasternack MS, Verret CR, Liu MA, Eisen HN. 1986. Serine esterase in cytolytic T lymphocytes. Nature 322:740–43 Spaeny-Dekking EHA, Kamp AM, Froelich CJ, Hack CE. 2000. Extracellular granzyme A, complexed to proteoglycans, is protected against inactivation by protease inhibitors. Blood 95:1465–72 Shresta S, Graubert TA, Thomas DA, Raptis SZ, Ley TJ. 1999. Granzyme A initiates an alternative pathway for granulemediated apoptosis. Immunity 10:595– 605 Pasternack MS, Bleier KJ, McInerney TN. 1991. Granzyme A binding to target cell proteins. J. Biol. Chem. 266:14,703–8

365

197. Suidan HS, Bouvier J, Schaerer E, Stone SR, Monard D, Tschopp J. 1994. Granzyme A released upon stimulation of cytotoxic T lymphocytes activates the thrombin receptor on neuronal cells and astrocytes. Proc. Natl. Acad. Sci. USA 91:8112–16 198. Irmler M, Hertig S, MacDonald HR, Sadoul R, Becherer JD, Proudfoot A, Solari R, Tschopp J. 1995. Granzyme A is an interleukin 1b-converting enzyme. J. Exp. Med. 181:1917–22 199. Beresford PJ, Kam C-M, Powers JC, Lieberman J. 1997. Recombinant human granzyme A binds to two putative HLAassociated proteins and cleaves one of them. Proc. Natl. Acad. Sci. USA 94: 9285–90 200. Beresford PJ, Jaju M, Friedman RS, Yoon MJ, Lieberman J. 1998. A role for heat shock protein 27 in CTL-mediated cell death. J. Immunol. 161:161–67 201. Zhang D, Pasternack MS, Beresford PJ, Wagner L, Greenberg AH, Lieberman J. 2000. Induction of rapid histone degradation by the cytotoxic T lymphocyte protease granzyme A. J. Biol. Chem. 276:3683–90 202. Anel A, Gamen S, Alava MA, SchmittVerhulst A-M, Pi˜neiro A, Naval J. 1997. Inhibition of CPP32-like proteases prevents granzyme B- and Fas-, but not granzyme A-based cytotoxicity exerted by CTL clones. J. Immunol. 158:1999– 2006 203. Beresford PJ, Xia Z, Greenberg AH, Lieberman J. 1999. Granzyme A loading induces rapid cytolysis and a novel form of DNA damage independently of caspase activation. Immunity 10:585–94 204. Kam C-M, Hudig D, Powers JC. 2000. Granzymes (lymphocyte serine proteases): characterization with natural and synthetic substrates and inhibitors. Biochem. Biophys. Acta 1477:307–23 205. Masson D, Tschopp J. 1987. A family of serine esterases in lytic granules of cytolytic T lymphocytes. Cell 49:679–85

5 Feb 2002

12:52

Annu. Rev. Immunol. 2002.20:323-370. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

366

AR

AR152-12.tex

RUSSELL

¥

AR152-12.SGM

LaTeX2e(2001/05/10)

P1: GJC

LEY

206. Edwards KM, Kam C-M, Powers JC, Trapani JA. 1999. The human cytotoxic T cell granule serine protease granzyme H has chymotrypsin-like (chymase) activity and is taken up into cytoplasmic vesicles reminiscent of granzyme B-containing endosomes. J. Biol. Chem. 274:30468–73 207. Shresta S, Goda P, Wesselschmidt R, Ley TJ. 1997. Residual cytotoxicity and granzyme K expression in granzyme A deficient cytotoxic lymphocytes. J. Biol. Chem. 272:236–44 208. Pham CTN, MacIvor DM, Hug BA, Heusel JW, Ley TJ. 1996. Long range disruption of gene expression by a PGKneo cassette. Proc. Natl. Acad. Sci. USA 93:13,090–95 209. Hanson RD, Hohn PA, Popescu NC, Ley TJ. 1990. A cluster of hematopoietic serine protease genes is found on the same chromosomal band as the human a/d Tcell receptor locus. Proc. Natl. Acad. Sci. USA 87:960–63 210. Heusel JW, Hanson RD, Silverman GA, Ley TJ. 1991. Structure and expression of a cluster of human hematopoietic serine protease genes found on chromosome 14q11.2. J. Biol. Chem. 266:6152–58 211. Zimmer M, Medcalf RL, Fink TM, Mattmann C, Lichter P, Jenne DE. 1992. Three human elastase-like genes coordinately expressed in the myelomonocyte lineage are organized as a single genetic locus on 19pter. Proc. Natl. Acad. Sci. USA 89:8215–19 212. Pilat D, Fink T, Obermaier-Skrobanek B, Zimmer M, Wekerle H, Lichter P, Jenne DE. 1994. The human Met-ase gene (GzmM): structure, sequence, and close physical linkage to the serine protease gene cluster on 19p13.3. Genomics 24: 445–50 213. MacIvor DM, Pham CTN, Ley TJ. 1999. The 50 flanking region of the human granzyme H gene directs expression to T/Natural Killer cell progenitors and lymphokine-activated killer cells in transgenic mice. Blood 93:963–73

214. Lowin B, Beermann F, Schmidt A, Tschopp J. 1994. A null mutation in the perforin gene impairs cytolytic T lymphocyte- and natural killer cell-mediated cytotoxicity. Proc. Natl. Acad. Sci. USA 91:11,571–75 215. Kojima H, Shinohara N, Hanaoka S, Someya-Shirota Y, Takagaki Y, Ohno H, Saito T, Katayama T, Yagita H, Okumura K. 1994. Two distinct pathways of specific killing revealed by perforin mutant cytotoxic lymphocytes. Immunity 1:357–64 216. Walsh CM, Mehrdad M, Liu C-C, Ueda R, Kurahara CG, Christensen JL, Huang MTF, Young JD-E, Ahmed R, Clark WR. 1994. Immune function in mice lacking the perforin gene. Proc. Natl. Acad. Sci. USA 91:10,854–58 217. Baker MB, Altman NH, Podack ER, Levy RB. 1996. The role of cellmediated cytotoxicity in acute GVHD after MHC-matched allogeneic bone marrow transplantation in mice. J. Exp. Med. 183:2645–56 218. Baker MB, Riley RL, Podack ER, Levy RB. 1997. Graft-versus-host diseaseassociated lymphoid hypoplasia and B cell dysfunction is dependent upon donor T cell-mediated Fas-ligand function, but not perforin function. Proc. Natl. Acad. Sci. USA 94:1366–71 219. Braun MY, Lowin B, French L, AchaOrbea H, Tschopp J. 1996. Cytotoxic T cell deficient in both functional Fas ligand and perforin show residual cytolytic activity yet lose their capacity to induce lethal acute graft-versus-host disease. J. Exp. Med. 183:657–61 220. Selvaggi G, Inverardi L, Levy RB, Brendel MD, Spielman J, Mintz DH, Podack ER, Ricordi C. 1995. Abrogation of graftversus-host disease following allotransplantation of cytotoxically deficient bone marrow across major histocompatibility barriers. Ann. NY Acad. Sci. 770:339–44 221. Baker J, Verneris MR, Ito M, Shizuru JA, Negrin RS. 2001. Expansion of cytolytic CD8+ natural killer T cells with limited

5 Feb 2002

12:52

AR

AR152-12.tex

AR152-12.SGM

LaTeX2e(2001/05/10)

P1: GJC

LYMPHOCYTE-MEDIATED CYTOTOXICITY

Annu. Rev. Immunol. 2002.20:323-370. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

222.

223.

224.

225.

226.

227.

228.

capacity for graft-versus-host disease induction due to interferon γ production. Blood 97:2923–31 Tsukada N, Kobata T, Aizawa Y, Yagita H, Okumura K. 1999. Graft-versus-leukemia effect and graft-versus-host disease can be differentiated by cytotoxic mechanisms in a murine model of allogeneic bone marrow transplantation. Blood 93:2738–47 Pan L, Teshima T, Hill GR, Bungard D, Brinson YS, Reddy VS, Cooke KR, Ferrara JLM. 1999. Granulocyte colonystimulating factor-mobilized allogeneic stem cell transplantation maintains graft-versus-leukemia effects through a perforin-dependent pathway while preventing graft-versus-host disease. Blood 93:4071–78 Schmaltz C, Alpdogan O, Horndasch KJ, Muriglan SJ, Kappel BJ, Teshima T, Ferrara JLM, Burakoff SJ, van den Brink MRM. 2001. Differential use of Fas ligand and perforin cytotoxic pathways by donor T cells in graft-versus-host disease and graft-versus-leukemia effect. Blood 97:2886–95 Blazar BR, Taylor PA, Vallera DA. 1997. CD4+ and CD8+ T cells each can utilize a perforin-dependent pathway to mediate lethal graft-versus-host disease in major histocompatibility complex-disparate recipients. Transplantation 64:571–76 Jiang Z, Podack ER, Levy RB. 2001. MHC-mismatched allogeneic BMT using perforin and/or FasL double-defective CD4+ donor T cells: involvement of cytotoxic function by donor lymphocytes prior to GvHD pathogenesis. Blood 98:390–97 Martin PJ, Akatsuka Y, Hahne M, Sale G. 1998. Involvement of donor T-cell cytotoxic effector mechanisms in preventing allogeneic marrow graft rejection. Blood 92:2177–81 Badovinac VP, Tvinnereim AR, Harty JT. 2000. Regulation of antigen-specific CD8+ T cell homeostasis by perforin and interferon-γ . Science 290:1354–57

367

229. Spielman J, Lee RK, Podack ER. 1998. Perforin/fas-ligand double deficiency is associated with macrophage expansion and severe pancreatitis. J. Immunol. 161: 7063–70 230. Stepp SE, Dufourcq-Lagelouse R, Le Deist F, Bhawan S, Certain S, Mathew PA, Henter JL, Bennett M, Fischer A, de Saint Basile G, Kumar V. 1999. Perforin gene defects in familial hemophagocytic lymphohistiocytosis. Science 286:1957– 59 231. Ericson KG, Fadeel B, Nilsson-Ardnor S, Soderhall C, Samuelsson A, Janka G, Schneider M, Gurgey A, Yalman N, Revesz T, Egeler R, Jahnukainen K, Storm-Mathiesen I, Haraldsson A, Poole J, de Saint Basile G, Nordenskjold M, Henter J. 2001. Spectrum of perforin gene mutations in familial hemophagocytic lymphohistiocytosis. Am. J. Hum. Genet. 68:590–97 232. Graubert TA, Russell JH, Ley TJ. 1996. The role of granzyme B in murine models of acute graft-versus-host disease and graft rejection. Blood 87:1232–37 233. Ebnet K, Hausmann M, Lehmann-Grube F, M¨ullbacher A, Kopf M, Lamers M, Simon MM. 1995. Granzyme A-deficient mice retain potent cell-mediated cytotoxicity. EMBO J. 14:4230–39 234. Wilharm E, Tschopp J, Jenne DE. 1999. Biological activities of granzyme K are conserved in the mouse and account for residual Z-Lys-SBzl activity in granzyme A-deficient mice. FEBS Lett. 459:139– 42 235. M¨ullbacher A, Ebnet K, Blanden RV, Hla RT, Stehle T, Museteanu C, Simon MM. 1996. Granzyme A is critical for recovery of mice from infection with the natural cytopathic viral pathogen, ectromelia. Proc. Natl. Acad. Sci. USA 93:5783–87 236. M¨ullbacher A, Waring P, Hla RT, Tran T, Chin S, Stehle T, Museteanu C, Simon MM. 1999. Granzymes are the essential downstream effector molecules for the control of primary virus infections by

5 Feb 2002

12:52

368

237.

Annu. Rev. Immunol. 2002.20:323-370. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

238.

239.

240.

241.

242.

243.

AR

AR152-12.tex

RUSSELL

¥

AR152-12.SGM

LaTeX2e(2001/05/10)

P1: GJC

LEY

cytolytic leukocytes. Proc. Natl. Acad. Sci. USA 96:13,940–55 Pereira RA, Simon MM, Simmons A. 2000. Granzyme A, a noncytolytic component of CD8(+) cell granules, restricts the spread of herpes simplex virus in the peripheral nervous sytems of experimentally infected mice. J. Virol. 74:1029– 32 Simon MM, Hausmann M, Tran T, Ebnet K, Tschopp J, Thahla R, M¨ullbacher A. 1997. In vitro- and ex vivo-derived cytolytic leukocytes from granzyme A x B double knockout mice are defective in granule-mediated apoptosis but not lysis of target cells. J. Exp. Med. 186:1781– 86 Davis JE, Smyth MJ, Trapani JA. 2001. Granzyme A and B-deficient killer lymphocytes are defective in eliciting DNA fragmentation but retain potent in vivo anti-tumor capacity. Eur. J. Immunol. 31: 39–47 Wolters PJ, Pham CTN, Muilenburg DJ, Ley TJ, Caughey GH. 2001. Dipeptidyl peptidase I is essential for activation of mast cell chymases, but not tryptases, in mice. J. Biol. Chem. 276:18,551–56 Adkison AM, Pham CTN. 2001. An essential role for dipeptidyl peptidase I in the activation of neutrophil-associated serine proteases and antibody-mediated acute arthritis. J. Clin. Invest. Submitted Toomes C, James J, Wood AJ, Wu CL, McCormick D, Lench N, Hewitt M, Widmer R, Ghaffar KA, Pemberton M, Hussein IR, Temtamy SA, Davies R, Read AP, Sloan P, Dixon MJ, Thakker NS. 1999. Loss-of-function mutations in the cathepsin C gene result in peridontal disease and palmoplantar keratosis. Nat. Genet. 23:421–24 Hart TC, Hart PS, Bowden DW, Michalec MD, Callison SA, Walker SJ, Zhang Y, Firatli E. 1999. Mutations of the cathepsin C gene are responsible for Papillon-Lef`evre syndrome. J. Med. Genet. 36:881–87

244. Nakano A, Nomura K, Nakano H, Ono Y, LaForiga S, Pulkkinen L, Hashimoto I, Uitto J. 2001. Papillon-Lef`evre syndrome: mutations and polymorphisms in the cathepsin C gene. J. Invest. Dermatol. 116:339–43 245. Hart TC, Hart PS, Michalec MD, Zhang Y, Firatli E, Van Dyke TE, Stabholz A, Zlorogorski A, Shapira L, Soskolne WA. 2000. Haim-Munk syndrome and Papillon-Lef e` vre syndrome are allelic mutations in cathepsin C. J. Med. Genet. 37:88–94 246. Hart TC, Hart PS, Michalec MD, Zhang Y, Marazita ML, Cooper M, Yassin OM, Nusier M, Walker S. 2000. Localization of a gene for prepubertal periodontitis to chromosome 11q14 and identification of a cathepsin C gene mutation. J. Med. Genet. 37:95–101 247. Gorlin RJ, Sedano H, Anderson VE. 1964. The syndrome of palmar-plantar hyperkeratosis and premature periodontal destruction of the teeth. J. Pediatr. 65:895–908 248. Via CS, Nguyen P, Shustov A, Drappa J, Elkon KB. 1996. A major role for the Fas pathway in acute graft-versus-host disease. Immunology 157:5387–93 249. Hsieh MH, Korngold R. 2000. Differential use of FasL- and perforinmediated cytolytic mechanisms by T-cell subsets involved in graft-versus-myeloid leukemia responses. Blood 96:1047– 55 249a. Smyth MJ, Thia KYT, Street ERA, MacGregor D, Godfrey DI, Trapani JA. 2000. Perforin-mediated cytotoxicity is critical for surveillance of spontaneous lymphoma. J. Exp. Med. 192:755– 60 250. Sun J, Bird CH, Sutton V, McDonald L, Coughlin PB, DeJong TA, Trapani JA, Bird PI. 1996. A cytosolic granzyme B inhibitor related to the viral apoptotic regulator cytokine response modifier A is present in cytotoxic lymphocytes. J. Biol. Chem. 271:27,802–9

5 Feb 2002

12:52

AR

AR152-12.tex

AR152-12.SGM

LaTeX2e(2001/05/10)

P1: GJC

Annu. Rev. Immunol. 2002.20:323-370. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

LYMPHOCYTE-MEDIATED CYTOTOXICITY 251. Sun J, Ooms L, Bird CH, Sutton VR, Trapani JA, Bird PI. 1997. A new family of 10 murine ovalbumin serpins includes two homologs of proteinase inhibitor 8 and two homologs of the granzyme B inhibitor (proteinase inhibitor 9). J. Biol. Chem. 272:15,434–41 252. Bird CH, Sutton VR, Sun J, Hirst CE, Novak A, Kumar S, Trapani JA, Bird PI. 1998. Selective regulation of apoptosis: the cytotoxic lymphocyte serpin proteinase inhibitor 9 protects against granzyme B-mediated apoptosis without perturbing the Fas cell death pathway. Mol. Cell. Biol. 18:6387–98 253. Bird CH, Blink EJ, Hirst CE, Buzza MS, Steele PM, Sun J, Jans DA, Bird PI. 2001. Nucleocytoplasmic distribution of the ovalbumin serpin PI-9 requires a nonconventional nuclear import pathway and the export factor Crm1. Mol. Cell. Biol. 21:5396–407 253a. Medema JP, de Jong J, Peltenburg LTC, Verdegaal EME, Gorter A, Bres SA, Franken KLMC, Hahne M, Albar JP, Melief CJM, Offringa R. 2001. Blockade of the granzyme B/perforin pathway through overexpression of the serine protease inhibitor PI-9/SPI-6 constitutes a mechanism for immune escape by tumors. Proc. Natl. Acad. Sci. USA 98:11515–20 254. Quan LT, Caputo A, Bleackley RC, Pickup DJ, Salvesen GS. 1995. Granzyme B is inhibited by the cowpox virus serpin cytokine response modifier A. J. Biol. Chem. 270:10,377–79 255. Tewari M, Telford WG, Miller RA, Dixit VM. 1995. CrmA, a poxvirusencoded serpin, inhibits cytotoxic Tlymphocyte-mediated apoptosis. J. Biol. Chem. 270:22,705–8 256. Macen JL, Garner RS, Musy PY, Brooks MA, Turner PC, Moyer RW, McFadden G, Bleackley RC. 1996. Differential inhibition of the Fas- and granule-mediated cytolysis pathways by the orthpoxvirus cytokine response mod-

257.

258.

259.

260.

261.

262.

263.

264.

265.

369

ifier A/SPI-2 and SPI-1 protein. Proc. Natl. Acad. Sci. USA 93:9108–13 Zhou Q, Snipas S, Orth K, Muzio M, Dixit VM, Salvesen GS. 1997. Target protease specificity of the viral serpin CrmA. J. Biol. Chem. 272:7797–800 Sieg S, Xia L, Huang Y, Kaplan D. 1995. Specific inhibition of granzyme B by parainfluenza virus type 3. J. Virol. 69:3538–41 Andrade F, Bull HG, Thornberry NA, Ketner GW, Casciola-Rosen LA, Rosen A. 2001. Adenovirus L4-100K assembly protein is a granzyme B-substrate that potently inhibits granzyme B-mediated cell death. Immunity 14:751–61 Thiele DL, McGuire MJ, Lipsky PE. 1997. A selective inhibitor of dipeptidyl peptidase I impairs generation of CD8+ T cell cytotoxic effector function. J. Immunol. 158:5200–10 Rossi CP, McAllister A, Tanguy M, Kagi D, Brahic M. 1998. Theiler’s virus infection of perforin-deficient mice. J. Virol. 72:4515–19 Murray PD, McGavern DB, Lin X, Njenga MK, Leibowitz J, Pease LR, Rodriguez M. 1998. Perforin-dependent neurologic injury in a viral model of multiple sclerosis. J. Neurosci. 18:7306– 14 Tang Y, Hugin AW, Giese NA, Gabriele L, Chattopadhyay SK, Fredrickson TN, Kagi D, Hartley JW, Morse HC. 1997. Control of immunodeficiency and lymphoproliferation in mouse AIDS: studies of mice deficient in CD8+ T cells or perforin. J. Virol. 71:1808–13 Riera L, Gariglio M, Valente G, Mullbacher A, Museteanu C, Landolfo S, Simon MM. 2000. Murine cytomegalovirus replication in salivary glands is controlled by both perforin and granzymes during acute infection. Eur. J. Immunol. 30:1350–55 Fleck M, Kern ER, Zhou T, Podlech J, Wintersberger W, Edwards CK 3rd, Mountz JD. 1998. Apoptosis mediated

5 Feb 2002

12:52

370

AR

AR152-12.tex

RUSSELL

¥

AR152-12.SGM

P1: GJC

LEY

by Fas but not tumor necrosis factor receptor 1 prevents chronic disease in mice infected with murine cytomegalovirus. J. Clin. Invest. 102:1431–43 266. Usherwood EJ, Brooks JW, Sarawar SR,

Annu. Rev. Immunol. 2002.20:323-370. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

LaTeX2e(2001/05/10)

Cardin RD, Young WD, Allen DJ, Doherty PC, Nash AA. 1997. Immunological control of murine gammaherpesvirus infection is independent of perforin. J. Gen. Virol. 78:2025–30

P1: FRK

February 1, 2002

12:47

Annual Reviews

AR152-FM

Annual Review of Immunology Volume 20, 2002

Annu. Rev. Immunol. 2002.20:323-370. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

CONTENTS FRONTISPIECE, Charles A. Janeway, Jr. A TRIP THROUGH MY LIFE WITH AN IMMUNOLOGICAL THEME, Charles A. Janeway, Jr.

xiv 1

THE B7 FAMILY OF LIGANDS AND ITS RECEPTORS: NEW PATHWAYS FOR COSTIMULATION AND INHIBITION OF IMMUNE RESPONSES, Beatriz M. Carreno and Mary Collins

29

MAP KINASES IN THE IMMUNE RESPONSE, Chen Dong, Roger J. Davis, and Richard A. Flavell

PROSPECTS FOR VACCINE PROTECTION AGAINST HIV-1 INFECTION AND AIDS, Norman L. Letvin, Dan H. Barouch, and David C. Montefiori T CELL RESPONSE IN EXPERIMENTAL AUTOIMMUNE ENCEPHALOMYELITIS (EAE): ROLE OF SELF AND CROSS-REACTIVE ANTIGENS IN SHAPING, TUNING, AND REGULATING THE AUTOPATHOGENIC T CELL REPERTOIRE, Vijay K. Kuchroo, Ana C. Anderson, Hanspeter Waldner, Markus Munder, Estelle Bettelli, and Lindsay B. Nicholson

55 73

101

NEUROENDOCRINE REGULATION OF IMMUNITY, Jeanette I. Webster, Leonardo Tonelli, and Esther M. Sternberg

125

MOLECULAR MECHANISM OF CLASS SWITCH RECOMBINATION: LINKAGE WITH SOMATIC HYPERMUTATION, Tasuku Honjo, Kazuo Kinoshita, and Masamichi Muramatsu

165

INNATE IMMUNE RECOGNITION, Charles A. Janeway, Jr. and Ruslan Medzhitov

KIR: DIVERSE, RAPIDLY EVOLVING RECEPTORS OF INNATE AND ADAPTIVE IMMUNITY, Carlos Vilches and Peter Parham ORIGINS AND FUNCTIONS OF B-1 CELLS WITH NOTES ON THE ROLE OF CD5, Robert Berland and Henry H. Wortis E PROTEIN FUNCTION IN LYMPHOCYTE DEVELOPMENT, Melanie W. Quong, William J. Romanow, and Cornelis Murre

197 217 253 301

LYMPHOCYTE-MEDIATED CYTOTOXICITY, John H. Russell and Timothy J. Ley

SIGNAL TRANSDUCTION MEDIATED BY THE T CELL ANTIGEN x

323

P1: FRK

Annu. Rev. Immunol. 2002.20:323-370. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

February 1, 2002

12:47

Annual Reviews

AR152-FM

CONTENTS

xi

RECEPTOR: THE ROLE OF ADAPTER PROTEINS, Lawrence E. Samelson INTERACTION OF HEAT SHOCK PROTEINS WITH PEPTIDES AND ANTIGEN PRESENTING CELLS: CHAPERONING OF THE INNATE AND ADAPTIVE IMMUNE RESPONSES, Pramod Srivastava CHROMATIN STRUCTURE AND GENE REGULATION IN THE IMMUNE SYSTEM, Stephen T. Smale and Amanda G. Fisher PRODUCING NATURE’S GENE-CHIPS: THE GENERATION OF PEPTIDES FOR DISPLAY BY MHC CLASS I MOLECULES, Nilabh Shastri, Susan

371

Schwab, and Thomas Serwold

395 427

463

THE IMMUNOLOGY OF MUCOSAL MODELS OF INFLAMMATION, Warren Strober, Ivan J. Fuss, and Richard S. Blumberg T CELL MEMORY, Jonathan Sprent and Charles D. Surh

GENETIC DISSECTION OF IMMUNITY TO MYCOBACTERIA: THE HUMAN MODEL, Jean-Laurent Casanova and Laurent Abel ANTIGEN PRESENTATION AND T CELL STIMULATION BY DENDRITIC CELLS, Pierre Guermonprez, Jenny Valladeau, Laurence Zitvogel, Clotilde Th´ery, and Sebastian Amigorena

495 551 581

621

NEGATIVE REGULATION OF IMMUNORECEPTOR SIGNALING, Andr´e Veillette, Sylvain Latour, and Dominique Davidson

669

CPG MOTIFS IN BACTERIAL DNA AND THEIR IMMUNE EFFECTS, Arthur M. Krieg

PROTEIN KINASE Cθ

709 IN

T CELL ACTIVATION, Noah Isakov and Amnon

Altman

RANK-L AND RANK: T CELLS, BONE LOSS, AND MAMMALIAN EVOLUTION, Lars E. Theill, William J. Boyle, and Josef M. Penninger PHAGOCYTOSIS OF MICROBES: COMPLEXITY IN ACTION, David M. Underhill and Adrian Ozinsky

761 795 825

STRUCTURE AND FUNCTION OF NATURAL KILLER CELL RECEPTORS: MULTIPLE MOLECULAR SOLUTIONS TO SELF, NONSELF DISCRIMINATION, Kannan Natarajan, Nazzareno Dimasi, Jian Wang, Roy A. Mariuzza, and David H. Margulies

853

INDEXES Subject Index Cumulative Index of Contributing Authors, Volumes 1–20 Cumulative Index of Chapter Titles, Volumes 1–20

ERRATA An online log of corrections to Annual Review of Immunology chapters (if any, 1997 to the present) may be found at http://immunol.annualreviews.org/

887 915 925

8 Feb 2002

9:55

AR

AR152-13.tex

AR152-13.SGM

LaTeX2e(2001/05/10) P1: GJC 10.1146/annurev.immunol.20.092601.111357

Annu. Rev. Immunol. 2002. 20:371–94 DOI: 10.1146/annurev.immunol.20.092601.111357

SIGNAL TRANSDUCTION MEDIATED BY THE T CELL ANTIGEN RECEPTOR: The Role of Annu. Rev. Immunol. 2002.20:371-394. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

Adapter Proteins∗

Lawrence E. Samelson Laboratory of Cellular and Molecular Biology, Center for Cancer Research, National Cancer Institute, 37 Convent Drive, Building 37, Room 1E24, Bethesda, Maryland, 20892-4255; e-mail: [email protected]

Key Words signaling pathways, linker for activation of T cells (LAT), plasma membrane microdomains, cytoskeleton, protein tyrosine kinases, tyrosine kinase substrates ■ Abstract Engagement of the T cell antigen receptor (TCR) leads to a complex series of molecular changes at the plasma membrane, in the cytoplasm, and at the nucleus that lead ultimately to T cell effector function. Activation at the TCR of a set of protein tyrosine kinases (PTKs) is an early event in this process. This chapter reviews some of the critical substrates of these PTKs, the adapter proteins that, following phosphorylation on tyrosine residues, serve as binding sites for many of the critical effector enzymes and other adapter proteins required for T cell activation. The role of these adapters in binding various proteins, the interaction of adapters with plasma membrane microdomains, and the function of adapter proteins in control of the cytoskeleton are discussed.

INTRODUCTION Signal transduction refers to the process by which extracellular events or cues are transmitted via a receptor or multiple receptors to the interior of the cell. Many of the current principles in the study of signal transduction have arisen from the study of various growth factor receptors. In these systems a transmembrane receptor binds ligand and then undergoes a conformational change or aggregation that has intracellular consequences. In many systems the receptor either is itself a protein kinase or is linked to one (1). The change in the receptor induced by binding results in kinase activation. The protein kinase(s) then phosphorylates a number ∗ The US Government has the right to retain a nonexclusive, royalty-free license in and to any copyright covering this paper.

371

8 Feb 2002

10:30

Annu. Rev. Immunol. 2002.20:371-394. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

372

AR

AR152-13.tex

AR152-13.SGM

LaTeX2e(2001/05/10)

P1: GJC

SAMELSON

of proteins, some of which may be effector enzymes, and this phosphorylation may result in their activation. In many cases these critical kinase substrates also include adapter molecules (2, 3). Adapter or linker molecules are proteins lacking enzymatic activity and are comprised of multiple binding domains and sequence motifs to which such domains bind. Phosphorylation of adapter molecules alters the surface of these molecules and allows additional enzymes or adapters to bind, which results in formation of multiprotein complexes. The associated and activated enzymes can be at the proximal end of a series of subsequent activation events, and multiple enzymes and enzyme pathways can be involved. The general theme of these processes is the transfer of information (4), that is, the transmission of an event on the exterior of the cell (receptor engagement) to activation and regulation of multiple intracellular events occurring at the plasma membrane, cytosol, and nucleus. Several principles leading to development of the above scheme represent some of the greatest advances in the field of signal transduction over the past decades. First, receptor molecules have both extracellular and intracellular functions. Ligands engage specific extracellular domains, thereby inducing conformational changes or aggregation that are transmitted intracellularly. The cytoplasmic component of growth factor receptors was recognized to contain multiple sites to which various signaling molecules could bind. In many cases the cytosolic tails of these receptors contain tyrosine residues, which upon phosphorylation serve as docking sites for proteins containing specific phosphotyrosine recognition domains such as SH2 domains. The second and related theme is that most signaling molecules are modular. Many enzymes and a large variety of adapter proteins contain domains for phosphotyrosine, polyproline, or lipid interaction as well as motifs to which some of these domains bind (Figure 1). There is great variety in the way in which these domains are brought together in a form of combinatorial diversity. Over the past decade adapters have been increasingly recognized throughout biology as critical to cell function and, in particular, cellular signaling. Investigation of adapter molecules in cells of the immune system has also been intense. Numerous review articles contain surveys of the multiple adapter proteins found in lymphocytes (5–11). Many of these molecules are well studied, while others have only been recently described. Instead of providing a catalog of the multiple adapter molecules isolated from lymphocytes, the goal of this article is a combination of the general and the specific. The first intent is to put the study of adapter molecules in a broad context by reviewing certain issues common to various signaling systems and to relate these to the signaling pathways coupled to the T cell antigen receptor (TCR). The bulk of the discussion, in the second part of the review, focuses on certain topics now actively studied by those interested in the role of adapters in TCR-mediated activation. One particular adapter, LAT (linker for activation of T cells), is extensively discussed throughout the review.

1 Feb 2002

10:21

AR

AR152-13.tex

AR152-13.SGM

LaTeX2e(2001/05/10)

P1: GJC

TCR SIGNALING AND ADAPTER PROTEINS

373

Annu. Rev. Immunol. 2002.20:371-394. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

SIGNALING VIA THE T CELL ANTIGEN RECEPTOR—BACKGROUND The most obvious difference between growth factor receptors and the TCR is the added complexity of the latter both in terms of receptor structure and in the molecules most closely coupled to the receptor. The ligand recognition component of the TCR, the α and β chains, which engage the complex of peptide and MHC molecule, have been extensively studied for over two decades (12, 13). Their complexity lies in the genetic and cellular mechanisms involved in creating millions of different clonotypic receptors in any individual. A discussion of these processes— genetic rearrangement, the pairing of chains to create stable dimers, and then the intricate intrathymic selection of receptors—is outside the scope of this paper. The recognition components interact with the nonpolymorphic CD3, γ , δ, and ε, and the TCRζ chain dimer (14). These molecules are integral TCR components and as such are required for TCR expression. The cytosolic components of these molecules contain a unique motif, the ITAM (immunoreceptor-based tyrosine activation motif), which has a consensus sequence of YxxI/L x(6-8)YxxI/L (15, 16). Each of the CD3 chains contains one such motif, whereas each TCRζ chain contains three. The actual arrangement and stoichiometry of CD3 and TCRζ chains within a TCR are unknown. However, for a working model of the TCR, one can consider each αβ to associate with a δε dimer, a γ ε dimer, and a TCRζ dimer. Each TCR in this configuration would thus contain ten ITAMs. ITAMs are necessary and sufficient for TCR-mediated activation (17, 18). The tyrosine residues within each ITAM become rapidly phosphorylated upon optimal TCR engagement. The phosphorylated ITAMs become subject to binding by additional molecules. In this fashion the CD3 and TCRζ chains themselves behave like adapter proteins containing motifs that are modifiable by phosphorylation and serve as binding sites for critical proteins. ITAM phosphorylation is mediated by two members of the Src family of PTKs found in T cells. Lck is the predominant enzyme involved in ITAM phosphorylation, while Fyn also has this capacity (19, 20). The most important consequence of ITAM tyrosine phosphorylation is the binding of ZAP-70, a member of a second family of PTKs involved in TCR signaling (21). The two phosphorylated tyrosines of each ITAM are bound by the tandem SH2 domains of ZAP-70 in a highly specific and cooperative fashion (22, 23). ZAP-70, once bound to the TCR in this fashion is activated by phosphorylation of the kinase domain activation loop mediated by the Src PTKs (24, 25). Other phosphorylations of ZAP-70 allow additional proteins to bind, giving ZAP-70 itself the role of a scaffold (26). The activated TCR is thus characterized by phosphorylated ITAMs associated with phosphorylated, activated ZAP-70. These activated PTKs then phosphorylate a large number of protein substrates. Over the past decade these proteins were first identified by detection of the subset of proteins phosphorylated on tyrosine residues after TCR engagement. Some of these proteins, such as the enzyme phospholipase Cγ 1 (PLC-γ 1), were known from

8 Feb 2002

10:31

Annu. Rev. Immunol. 2002.20:371-394. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

374

AR

AR152-13.tex

AR152-13.SGM

LaTeX2e(2001/05/10)

P1: GJC

SAMELSON

other systems and were rapidly identified (27, 28). Others were isolated because of their phosphorylation. These include other enzymes and, importantly, several of the adapter molecules such as LAT and SLP-76, discussed extensively in this review. The recognition that adapter molecules, both those subject to phosphorylation and others that are not, are critical in TCR signaling creates a number of analogies to other signaling systems in which adapters are of central importance in the signal transduction process (Figure 2). Among the most prominent PTK substrates detected after TCR engagement is a protein of 36–38 kDa (29). Early studies demonstrated that this protein could be detected in a complex with a number of other PTK substrates including PLC-γ 1 and the small linker molecule Grb2 (see below) (30–32). Despite both the ease of detection and its association with known proteins, the protein known then as pp36 proved difficult to isolate. Ultimately modified protein purification conditions enabled Zhang et al. to obtain amino acid sequence and clone the cDNA encoding this protein, which was named LAT (linker for activation of T cells) (33). Sequence analysis demonstrated that LAT is a member of a relatively unusual class of transmembrane adapter molecules. It is a class III–type protein, lacking a signal sequence. It contains a short extracellular sequence, a transmembrane domain, and a long cytosolic component containing nine tyrosine residues conserved between mouse and human LAT. Early studies also revealed that two cysteine residues (C26 and C29) are subject to posttranslational palmitoylation, which is responsible for specific localization within the plasma membrane (see below) (34). The central role of LAT in TCR-mediated signaling has been revealed in studies of the Jurkat T cell and Jurkat variants that lack LAT (35, 36). Jurkat cells, activated by cross-linking the TCR with monoclonal antibodies directed at either TCRβ or CD3ε show activation of multiple intracellular biochemical pathways and transcriptional elements leading to induction of interleukin 2 synthesis. These events include calcium elevation and ERK, AP-1, and NFAT activation. They do not occur in the Jurkat variants that lack LAT, but all activation events can be observed if LAT expression is restored following transfection of LAT cDNA. LAT function is also required for intrathymic development of normal T cells (37). In animals that are genetically modified to lack the LAT gene, intrathymic development of T cells is blocked at an early stage, and thus no T cells are found in the lymph nodes or spleens. LAT is rapidly phosphorylated on tyrosine residues following TCR engagement. Overexpression studies in fibroblastoid lines revealed that ZAP-70 and Syk are the PTKs most likely to be responsible for these phosphorylations. This conclusion is supported by the observation that LAT is very poorly phosphorylated in T cells lacking ZAP-70 (38). Nonetheless low levels of LAT phosphorylation are induced by activated forms of Lck, and it is possible that some of the multiple tyrosines in the LAT can be phosphorylated by PTKs other than ZAP-70. The consequence of these multiple phosphorylations on LAT in T cells is that a number of signaling proteins bind at these sites following TCR engagement.

1 Feb 2002

10:21

AR

AR152-13.tex

AR152-13.SGM

LaTeX2e(2001/05/10)

P1: GJC

Annu. Rev. Immunol. 2002.20:371-394. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

TCR SIGNALING AND ADAPTER PROTEINS

375

The proteins that interact with LAT fit into two categories, enzymes and adapter proteins. A full description of all these proteins is outside the scope of this manuscript. It is also likely that additional proteins that bind LAT remain to be described. One set of LAT-binding adapter proteins, the Grb2 family, contains Grb2, Grap, and Gads. All are of very simple design, consisting of a central SH2 domain flanked by two SH3 domains. These domains show considerable sequence homology between the three proteins. The Gads protein, known also by a series of other names [Grpl, MONA, Grf40, reviewed in (9)] also contains an internal proline-rich region of unknown function between the SH2 and the C-terminal SH3 domains (39). These simple linker proteins bind to tyrosine phosphorylated LAT via their SH2 domains. LAT structure-function studies have shown that three distal LAT tyrosine residues, Tyr 171, 191, and 226 (human sequence), following phosphorylation bind Grb2, while Gads binds only phosphorylated 171 and 191 (40). Grb2, Grap, and Gads are bifunctional molecules in the sense that their SH2 domains bind one protein, in this case phosphorylated LAT, while their SH3 domains bind other proteins, which contain the relevant proline-rich sequences. Grb2 and the related proteins are cytosolic and their binding to phosphorylated LAT translocates them and their SH3-associated proteins to a different cellular compartment, the plasma membrane. Thus, a consequence of LAT phosphorylation is the association of a number of Grb2-binding proteins at the membrane. Grb2 is a ubiquitously expressed protein, and its function in many systems is to shuttle associated signaling proteins to tyrosine phosphorylated receptors or adapter molecules following ligand activation (41). The list of proteins that bind Grb2 SH3 domains is vast. In T cells several proteins have been prominently described (42). SOS is well known from many studies in nonlymphoid tissues as a critical activator of the small G protein Ras (43). Grb2-mediated translocation of SOS from the cytosol to the plasma membrane brings it to the site of Ras localization. The Grb2-SOS complex has been detected in T cells, and SOS has been found in association with LAT following TCR engagement (33, 44). Recently there has been the identification of another Ras activator, in T cells, Ras-GRP, and thus the relative importance and function of each of these effector molecules remains unclear (45, 46). Cbl is another Grb2-associated protein found in T cells, bound to phosphorylated LAT. Cbl too has many domains and interactions. Recently Cbl has been shown to be part of a ubiquitin ligase assembly, and its function in that capacity in T cells also needs much additional investigation (47). Another small linker molecule, Shc, was originally demonstrated to bind to growth factor receptors and simultaneously to Grb2, which in turn in these studies was bound to SOS. Thus, in these systems an additional molecule was interposed between the receptor and the effector enzyme, SOS. Shc interactions with Grb2-SOS complexes have been studied in T cells (48). Another member of this simple adapter family is Grap, which also binds LAT following LAT phosphorylation (49). Grap-associated proteins include SOS, dynamin, and Sam68. Currently there is no good explanation for why T cells need two

1 Feb 2002

10:21

Annu. Rev. Immunol. 2002.20:371-394. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

376

AR

AR152-13.tex

AR152-13.SGM

LaTeX2e(2001/05/10)

P1: GJC

SAMELSON

adapter proteins, Grb2 and Grap, with similar SH3 binding specificities. The third member of the family is Gads, which differs from Grb2 in its tissue distribution. While Grb2 is ubiquitously expressed, Gads is found only in hematopoietic cells. Gads also has unique SH3 specificity. While Grb2 does not bind to SLP-76 in T cells, Gads specifically interacts with SLP-76. Thus, Gads brings another critical linker molecule to phosphorylated LAT (50). Recently Gads has been shown to interact with the serine-threonine kinase HPK (51). Certainly, the full inventory of Gads-associated proteins remains to be described. SLP-76 was first identified as a tyrosine kinase substrate that could be affinity purified in vitro by binding to Grb2 (52). It is a 76-kDa protein expressed exclusively in hematopoietic cells. SLP-76 lacks enzymatic activity and structurally can be divided into three domains. At the amino-terminal end, multiple tyrosines become phosphorylated on TCR engagement. The central domain is rich in proline residues including those that interact with the SH3 domains of Gads. The carboxy-terminal end of the protein contains an SH2 domain. Overexpression of SLP-76 in Jurkat cells led to an increase in TCR-mediated activation of NFAT and IL-2 promoters (53). No increase in calcium mobilization was seen in such experiments, though ERK activation was enhanced. All three SLP-76 domains are required for this augmentation of function (54). Study of a Jurkat mutant lacking SLP-76 revealed that following TCR engagement these cells show decreased calcium flux and no ERK activation (55). Not unexpectedly in these cells, IL-2 promoter activity was not increased in response to TCR cross-linking. SLP-76 also has a critical role in normal thymocyte development. Mice lacking SLP-76 fail to generate normal, peripheral T cells (56, 57). An intrathymic block occurs in T cell development at an early stage. The T cell phenotype of these mice resembles that of mice that lack the LAT gene, as described above. Both of these mice demonstrate the same developmental block and absence of mature T cells. SLP-76 functions as an adapter protein that binds multiple effector molecules, which can be brought to LAT via its association with Gads (50). A recent study demonstrated that a LAT-SLP-76 chimera containing only the transmembrane domain and palmitoylation sites of LAT, i.e., the raft targeting region (see below), suffices to reconsitute LAT-deficient variants of Jurkat. Although overexpression of such constructs in Jurkat variants does not fully mimic all the complex interactions that might occur in more physiologic T cell systems, the result does emphasize the significance of SLP-76 recruitment (58). Upon TCR engagement, activated ZAP70 phosphorylates multiple tyrosine residues in the amino-terminal end of SLP76. These phosphorylated residues serve as binding sites for a series of proteins containing SH2 domains. These include Vav, a guanine nucleotide exchange factor for the G-proteins of the Rac family; Nck, itself an adapter protein that interacts with the serine-threonine kinase Pak1; and the PTK Itk (59–61). These associated molecules have been implicated in the regulation of a number of pathways critical to T cell activation. Vav and Nck may integrate the activation of a number of pathways involved in both gene transcription and cytoskeletal rearrangement as discussed below. The association of Itk with SLP-76 brings this PTK into close contact with

1 Feb 2002

10:21

AR

AR152-13.tex

AR152-13.SGM

LaTeX2e(2001/05/10)

P1: GJC

TCR SIGNALING AND ADAPTER PROTEINS

377

Annu. Rev. Immunol. 2002.20:371-394. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

PLC-γ . Recent studies suggest that Itk activation is required for optimal PLC-γ 1 phosphorylation and activation (62). SLP-76 SH2 domain interacts with a 130kDa protein named both SLAP (SLP-76-associated phosphoprotein) and Fyb (Fyn binding protein). This protein is also a multidomain adapter protein containing a proline-rich region, a tyrosine-rich region, and an SH3-like domain. The gene encoding this molecule has been genetically deleted by two groups (62a, 62b). The new proposed name for the protein is ADAP (adhesion and degranulation promoting adaptor protein).

LAT AND THE RAFT MODEL OF T CELL ACTIVATION The classic lipid bilayer model describing the molecular organization of the plasma membrane has been modified over the past decade because of the realization that the plasma membrane is not a homogenous array of glycerophospholipids (63–65). Instead there is considerable heterogeneity of lipids in the membrane. Glycosphingolipids and cholesterol were shown to self-associate in model membrane systems, and similar phenomena were then observed in plasma membranes isolated from cells. The aggregation of glycosphingolipids and cholesterol is thought to induce the formation of microdomains in the membrane that are distinct from the more abundant and diffuse glycerolipids. These domains are known by a large number of acronyms: GEMs (glycolipid enriched microdomains), DIGs (detergent insoluble glycolipid-enriched membranes, DRMs (detergent-resistant membranes), or rafts. Many of these names reflect the standard method used to isolate such domains, which is the inability of non-ionic detergents such as Triton X-100 to solubilize these domains from plasma membranes in the cold. Such insoluble material can be separated from solubilized cellular material by sucrose gradient centrifugation. These microdomains or rafts are also enriched in a number of molecules relevant to receptor-mediated signaling. These include the lipid substrates of PLC-γ and GPI (glycosylphosphatidylinositol)-anchored proteins, including such molecules expressed on T cells as Thy1 and Ly6. Additionally Ras, various G proteins, and members of the Src PTK family are enriched in these domains (66). For the proteins in rafts that are not GPI-anchored, one shared characteristic is posttranslational modification by several lipids. For example, Ras is both palmitoylated and farnesylated. Most of the Src PTKs (except for Src itself) are modified by myristoylation and palmitoylation (67, 68). Evidence obtained from study of the T cell–specific Src family PTK, Lck indicates that both of these lipid modifications are necessary for raft localization targeting and phosphorylation of the TCRζ chain (69). Studies on the FcεRI receptor were the first to demonstrate that immunoreceptor activation involves interactions with rafts. These investigators showed that this receptor associated rapidly with rafts upon activation (70, 71). The very early tyrosine phosphorylation of receptor subunits by the Src family PTK, Lyn depended on this localization. Several groups subsequently made similar observations about TCR activation in cell lines and thymocytes (72, 73). Receptor engagement increased

1 Feb 2002

10:21

Annu. Rev. Immunol. 2002.20:371-394. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

378

AR

AR152-13.tex

AR152-13.SGM

LaTeX2e(2001/05/10)

P1: GJC

SAMELSON

the level of TCR association with rafts, and TCR subunits demonstrated enhanced tyrosine phosphorylation. Phosphorylated, and thus activated, ZAP-70 was found associated with these activated receptors. All of these studies depend essentially on a negative result, the inability of detergent extraction of the receptor at certain conditions of activation, cell number, temperature, and detergent type. Despite these results, several investigators at the time did not observe enhanced TCR association with membrane rafts. A possible explanation for the conflicting results was proposed by investigators who took an independent approach to the question (74). They visualized rafts using fluorescent cholera toxin B subunit, which binds glycosphingolipids. Antibodies to the subunit induced cross-linking of these lipids forming patches that are visible microscopically. With this technique they were able to observe colocalization in rafts of Lck and TCR subunits. Interestingly TCR subunits patched in this fashion were more sensitive to detergent extraction than were other raft-associated proteins. The authors speculated that this result indicated that the TCR is more weakly associated with rafts than are other molecules, which might constitutively localize in these domains. The study of the LAT protein contributed to an understanding of raft localization and T cell activation. Examination of the LAT amino acid sequence revealed the presence of two cysteine residues (positions 26 and 29) adjacent to the putative transmembrane domain of the protein. Since juxtamembrane cysteines are likely to be targets of the membrane-associated palmitoylation machinery, T cells were labeled with [3H]-palmitate and incorporation was demonstrated (34). Mutation of these two cysteines to alanine blocked this incorporation. These mutations and especially mutation of Cys 26 alone also had dramatic effect on LAT localization to rafts, as determined by the standard biochemical extraction assay. Though mutation of the cysteine residues did not affect membrane localization of LAT, mutation of Cys 29, partially, and Cys 26 fully inhibited LAT localization to rafts. LAT with these mutations were also expressed in T cells and examined following TCR engagement. Mutation of cysteine 26 had a dramatic effect, and no LAT tyrosine phosphorylation was detected. From these studies it was concluded that LAT had to be in rafts for it to be phosphorylated. The majority of LAT molecules localize in rafts as determined by the sucrose gradient analysis. As expected in samples from nonactivated T cells, essentially none of the LAT substrates are detected in these fractions. However, upon activation and LAT phosphorylation, one observes a translocation of LAT binding proteins to the raft fraction. The fraction of such LAT binding proteins as PLC-γ , Cbl, or Grb2 that shifts in this way is small, but this is presumably the fraction that is functionally active. The consequences of the absence of LAT localization to rafts and the failure of LAT tyrosine phosphorylation were thought likely to be dramatic, and this prediction was confirmed in two subsequent studies (36, 75). In both, independently isolated Jurkat mutants lacking the LAT molecule were used. The absence of LAT had no effect on TCR-induced receptor phosphorylation or ZAP-70 activation, but all steps distal to this were inhibited. There was no activation of PLC-γ 1, and thus there was minimal calcium flux or ERK activation. Several

1 Feb 2002

10:21

AR

AR152-13.tex

AR152-13.SGM

LaTeX2e(2001/05/10)

P1: GJC

Annu. Rev. Immunol. 2002.20:371-394. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

TCR SIGNALING AND ADAPTER PROTEINS

379

transcription factors were not activated. In both of these studies, reintroduction of wild-type LAT restored all function. Introduction of LAT deficient in one residue, Cys 26, had the same effect as though no LAT was present in the cell. Thus, without LAT localization to rafts, there is no LAT tyrosine phosphorylation, no translocation of critical signaling molecules to the rafts, and no T cell activation. If LAT is in membrane rafts and the TCR in the resting state is not, how do these structures interact, and how does the TCR become raft-associated, albeit weakly? It is possible that receptor engagement alters some biophysical properties of the TCR. A common model for TCR activation requires some level of TCR-TCR interaction, which at the most extreme would be TCR oligomerization or aggregation (76). Hypothetically this process could expose different regions of TCR subunits, which might enhance interactions with different membrane components. Similarly, receptor aggregates might also be more likely to trap rafts. Interactions of other TCR and LAT associated proteins have been described. Some or all of these interactions might be involved in bringing the activated TCR and associated PTKs to LAT molecules. The Lck PTK is located in rafts, and its SH2 domain can bind a phosphorylated tyrosine residue in the activated ZAP-70 PTK (77, 78). This intermolecular bridge may bring TCRs bearing activated ZAP-70 to rafts. Additionally, a subset of CD4 interacts with LAT (79). Other CD4 molecules may interact with Lck. Since CD4 interacts with MHC class II molecules and since some CD4 molecules exist as dimers (80), one can construct a model by which TCR and CD4 engage the same MHC, and the CD4-associated Lck and LAT, both in rafts, are brought to the TCR. Several additional molecules have been proposed to bind both TCR and LAT, thus potentially linking these molecules. PLC-γ 1 contains two SH2 domains, and it is well documented that the N-SH2 domain interacts with LAT. Williams et al., studying SH2 fusion proteins, proposed that the C-SH2 domain interacts with phosphorylated residues on activated ZAP-70 in a fashion similar to that described above for Lck (81). Deckert et al. have proposed that a molecule originally isolated as an Abl-SH3 interacting protein, 3BP2, can interact via its SH2 domain with both ZAP-70 and LAT (82). Though this interaction cannot be simultaneous because 3BP2 has only one SH2 domain, perhaps the molecule could multimerize. The authors of this study suggest that a functional coupling could occur, leading to the observed enhancement in T cell activation. Finally a small adapter protein known as Shb does contain two separate phosphotyrosine-binding domains (83). A classic SH2 domain in Shb was shown to bind TCRζ chain on phosphorylated tyrosine residues. A non-SH2 phosphotyrosine binding domain bound phosphorylated LAT. Expression of a mutant form of Shb with a defective SH2 domain inhibited LAT phosphorylation and distal signaling events. These data suggest that Shb links between the TCR and LAT have major functional significance. A recent study provides further insight into TCR-LAT interactions. Harder & Kuhn incubated T cells with anti-TCR antibodies coupled to beads (84). Activation via the TCR was induced with warming to 37◦ C, and the cells were subjected to nitrogen cavitation. The material that associated with the TCR on the bead over

1 Feb 2002

10:21

Annu. Rev. Immunol. 2002.20:371-394. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

380

AR

AR152-13.tex

AR152-13.SGM

LaTeX2e(2001/05/10)

P1: GJC

SAMELSON

different times of warming was then analyzed biochemically. No detergent was used in these preparations, and what was isolated, presumably, were membrane patches. In addition to the TCR subunits and the associated ZAP-70, the investigators showed that over time LAT was detected in these complexes. Some of the proteins that bind activated LAT such as PLC-γ 1, Grb2, and Cbl were also found in these patches. In contrast what they did not see associated with the TCR were the raft-associated PTKs, Lck or Fyn. The association of TCR and LAT in these membrane fragments depended on tyrosine kinase activity, and mutant LAT, which lacked the cysteines required for palmitoylation, did not co-isolate with the TCR. The conclusion from these studies is that TCR-LAT interactions do not represent an interaction of the receptor with LAT via large lipid aggregations in which Lck and Fyn would also be expected to be found. Instead the study supports the idea that protein-protein interactions induced by TCR activation and dependent on tyrosine phosphorylation control the critical TCR-LAT association. Interest in the raft model of immunoreceptor signaling has been intense, and numerous studies have expanded an understanding of the interactions between signaling receptors and the plasma membrane. However, it is wise to recognize that there are still a number of problems with this model that investigators in this field must acknowledge. The main concern is methodologic. Nearly all studies of raft function rely on the crudest of preparations, material that fails to be solubilized by certain detergents. Moreover, the preparations are made from cells that have been chilled to near freezing temperature. In so doing one may force interactions of proteins and lipids that might not normally occur under physiologic conditions. Efforts to reproduce raft isolation without detergent or chilling are rare, but, as described above, some investigators have begun this process. A similar criticism can be made about methodologies in which raft components are visualized after heavy cross-linking induced by multivalent toxins and antibodies. These studies demonstrate clustering of molecules shown to colocalize by biochemical analysis, but clearly the system is being forced. Ideally, imaging techniques could be used to demonstrate membrane heterogeneity. Investigators have attempted to use fluorescence energy transfer (FRET) techniques to visually demonstrate clustering of GPI-linked proteins. Two groups have reported contradictory results, with one obtaining results consistent with microdomains of less than 70 nM containing just a few molecules, whereas the other group saw no such structures (85, 86). The discrepancy could be resolved if only a few molecules were clustered over a minor fraction of the surface, or if such structures were short-lived. There is certainly strong evidence that membrane heterogeneity exists in model systems and in cells. There is much evidence that membrane microdomains are relevant for signaling in lymphocytes. The dramatic effect on T cell signaling of the cysteine LAT mutations is an example of such an experiment. Frequently however one senses that rafts have come to imply long-lived, well-defined membrane substructures. The conclusion that membrane heterogeneity is a dynamic process and that transient interactions of lipids and proteins are likely is much more reasonable and cautious.

1 Feb 2002

10:21

AR

AR152-13.tex

AR152-13.SGM

LaTeX2e(2001/05/10)

P1: GJC

TCR SIGNALING AND ADAPTER PROTEINS

381

Annu. Rev. Immunol. 2002.20:371-394. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

SIGNALING COMPLEXES: THE LAT-PHOSPHOLIPASE Cγ 1 EXAMPLE PLC-γ 1 was one of the first enzymes demonstrated to be a PTK substrate in T lymphocytes, following its identification as a PTK substrate in growth factor receptor tyrosine kinase systems. Phosphorylation of PLC-γ 1 on multiple tyrosine residues is required for its activation (87). PLC-γ 1 is a central signaling molecule in T cells as well as other cells activated by PTKs. Activation of this enzyme leads to hydrolysis of phosphatidyl inositol (4,5)-bisphosphate to inositol (1,4,5)-trisphosphate and diacylglycerol (88). The former regulates intracellular calcium mobilization, and the latter regulates protein kinase C activation. Recent studies in T cells demonstrate that calcium and diacylglycerol regulate RasGRP, a newly described activator of Ras. By contributing to Ras activation, PLC-γ 1 thus indirectly can control PI3 kinase and MAP kinase cascades (45, 46). The multiple binding interactions that engage PLC-γ 1 and molecules that regulate this enzyme are now under intensive scrutiny, and the results of these studies serve as an excellent example of signaling complexes containing adapter molecules involved in TCR activation. An interaction of PLC-γ 1 and a 36–38 kDa protein was described long before LAT was isolated and characterized. The association was seen after T cell activation and was dependent on the two PLC-γ 1 SH2 domains (30, 89). This and a subsequent study showed that the N-terminal SH2 domain was more specific for phospho-LAT, but in the later study the C-terminal SH2 domain was shown to bind LAT as well as other proteins (90). Mutation of the N-SH2 resulted in depressed tyrosine phosphorylation of PLC-γ 1 following TCR binding. The sites of PLCγ 1 interaction with LAT were addressed in a LAT structure-function study. TCR engagement of LAT-deficient variants of the Jurkat cell line (J.CaM2) failed to activate PLC-γ 1 and thus failed to elevate intracellular calcium or activate ERK. Reconstitution with wild-type LAT restored these pathways. Zhang et al. created a series of stable lines in which J.CaM2 was reconstituted with LAT mutants containing one or more tyrosine-to-phenylalanine mutations (40). The residues adjacent to Tyr 132, YLVV, form a consensus binding sequence for PLC-γ 1 SH2 domains. Mutation of tyrosine at this site abrogated PLC-γ 1-LAT association and PLC-γ phosphorylation on tyrosine residues. Cell lines expressing this mutant showed altered calcium flux following TCR engagement. The rapid onset of calcium elevation was observed, but the sustained influx, normally seen following TCR engagement, did not occur. Two additional reports showed subtly different results. In one, the 132 mutations inhibited PLC-γ 1 phosphorylation and calcium flux (91). In the other, phosphorylation of PLC-γ 1 persisted, but calcium flux was sharply curtailed (92). The differences are likely due to slight variations in technique; nonetheless, all three studies point to the critical role of this site on LAT for PLC-γ 1 function. The three distal tyrosine residues of LAT (Tyr 171, 191, and 226) all are found within YXNX motifs. Phosphorylation of this motif defines sites for binding via the SH2 domains of Grb2 or related adapters. Interestingly, mutation of all three

1 Feb 2002

10:21

Annu. Rev. Immunol. 2002.20:371-394. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

382

AR

AR152-13.tex

AR152-13.SGM

LaTeX2e(2001/05/10)

P1: GJC

SAMELSON

of these sites also had a strong impact on PLC-γ 1 (40). The PLC-γ 1-LAT association was not detected; PLC-γ 1 tyrosine phosphorylation was nearly completely inhibited, and calcium flux was partially inhibited. These sites might bind the C-terminal SH2 domain of PLC-γ 1 directly or might bind PLC-γ 1 indirectly as described below. These three sites were also tested for binding by Grb2 and the related protein Gads. Mutation of any one of the three distal tyrosine residues (171, 191, or 226) had no effect on Grb2 or Gads binding, which suggests a degree of redundancy in the system. Loss of both 171 and 191 decreased Grb2 binding, and only mutation of all three of these tyrosines blocked Grb2 binding. Gads binding proved more restricted because mutation of both 171 and 191 inhibited interaction. Since a major binding partner of Gads is SLP-76, which in turn interacts with PLC-γ 1, these results account for the loss of SLP-76 binding in the double Tyr 171, 191 mutant. The multidomain adapter SLP-76 is critical to T cell activation, and as mentioned above, cell lines deficient in this molecule have a significant defect in PLC-γ 1 activation (55), a result that now can be explained. As noted above, many molecules bind to SLP-76 following phosphorylation of its N-tyrosines or via its C-terminal SH2 domain. Yablonski et al. demonstrated the significance of a proline rich region (157–223) that interacts with the PLC-γ 1 SH3 domain (93). This stretch of prolines is distinct from the residues involved in Gads binding (224– 265). An SLP-76-deficient variant of Jurkat was used as the recipient for SLP-76 mutants in a structure-function study. Constitutive association between PLC-γ 1 and SLP-76 was dependent on the PLC-γ 1 SH3 domain and an inducible increase in association that they attributed to direct and indirect interactions via LAT. They proposed that two previously defined complexes, LAT-Gads-SLP-76 and LATPLC-γ 1, in fact interact via the binding of SLP-76 to PLC-γ 1. They suggest, in other words, that a multiprotein complex nucleated at LAT contains Gads, SLP-76, and PLC-γ 1, and in this complex both Gads and PLC-γ 1 bind LAT. This conclusion is bolstered by another study in which they demonstrate that the functional complex of these molecules must be bound to the same LAT molecule (92). How the individual phosphorylation sites on all PTK substrates are targeted by various PTKs in T cells is still under investigation. Nonetheless, it is clear that members of the Tec PTK family are required for full PLC-γ 1 phosphorylation and activation. In support of this conclusion is the observation that deletion of the Tec PTK Itk or deletion of two Tec PTKs, Itk and Txk/Rlk, produces defects in sustained calcium elevation following TCR engagement (62, 94). Overexpression of Txk/Rlk in transgenic mice also showed enhanced PLC-γ 1 phosphorylation and calcium flux (95). If regulation of PLC-γ 1 activation at LAT is likely to involve Tec family PTKs, the next question is how these enzymes are targeted to this site. A full answer is not yet in hand, though the multiple domains of Itk interact with many molecules, and one or more of these interactions might be relevant to this question of targeting (61). The N-terminal PH domain of Itk or the palmitoylation of Txk/Rlk are likely to control plasma membrane localization. The TH (Tec homology) region contains a proline-rich region that interacts with a Grb2 SH3 domain. The Itk SH3 domain interacts with proline-rich regions

1 Feb 2002

10:21

AR

AR152-13.tex

AR152-13.SGM

LaTeX2e(2001/05/10)

P1: GJC

Annu. Rev. Immunol. 2002.20:371-394. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

TCR SIGNALING AND ADAPTER PROTEINS

383

of PLC-γ , though the authors of the study on PLC-γ 1-SLP76 interactions think that this interaction is not physiologically relevant (93). The Itk SH2 domain has been shown, by several investigators, to interact with SLP-76 (60, 61). Finally, Itk interactions with LAT have been reported, though it is not known whether this is a direct Itk SH2-mediated interaction or whether it is indirect (96). Thus, it is not now clear whether one predominant mode of Itk interaction with LATassociated molecules predominates, or whether there are multiple mechanisms of interaction. The consensus derived from many investigators is that a complex of LAT, Gads, SLP-76, PLC-γ 1, and a Tec PTK, usually Itk, regulates PLC-γ 1 activation in response to TCR signaling. These conclusions follow from characterization of cell lines and mice lacking expression of one of these proteins and from an extensive analysis of the fine specificity of multiple protein-protein interactions. This model has certain strikingly positive features, but it remains incomplete in a number of ways. Its greatest strength is the manner in which the strength of multiple protein-protein interactions is likely to be far greater than the sum of individual interactions. Ladbury and Arold have noted that the difference in affinity between specific and nonspecific SH2 and SH3-mediated individual interactions is usually less than two orders of magnitude (97). These authors note that the assembly of multimolecular complexes involving many such interactions ensures that a proper assembly must occur before signaling transpires. Such interactions can be defined as highly cooperative. Interactions of this sort are central to the generation of the multiprotein complex regulating PLC-γ 1. The multiple individual interactions might zip together the functional PLC-γ 1 machine. In addition such multistage events also offer great potential for regulation, as inhibition of any of the multiple steps might block assembly. However, before defining such a putative entity as a signalosome, a term that might imply a far more stable structure of defined stoichiometry, a number of caveats and concerns must be discussed. The major issue is that LAT has been shown to bind a large number of different signaling molecules. Gads, SLP-76, and PLC-γ 1, described above, are just a subset. Moreover, even these molecules have additional binding partners. Gads also binds the serine-threonine protein kinase HPK, and thus a different LAT-Gads complex may exist (51, 98). Similarly SLP-76, as described in detail below, has many additional partners. Competition for LAT can also occur. Grb2 binding to LAT is well defined, and Grb2 is capable of coordinating a number of LAT-based complexes including interactions with SOS and Cbl. The SH2 domains of Gads and Grb2 have similar binding characteristics, and both were shown to bind to two of the three distal tyrosine residues of LAT (171 and 191) following activation. It is not clear what the relative binding affinities between Gads and Grb2 are for these sites, nor has the relative stoichiometry of binding between LAT, Gads, and Grb2 been determined. Clearly, though, the issue of whether and how much Grb2 or Gads is bound to LAT would have a great impact on which other molecules are brought to LAT. It is unlikely that all possible interactions can occur at one LAT molecule at the same time because there are too many possibilities leading to competition at the same site. The competition for

1 Feb 2002

10:21

Annu. Rev. Immunol. 2002.20:371-394. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

384

AR

AR152-13.tex

AR152-13.SGM

LaTeX2e(2001/05/10)

P1: GJC

SAMELSON

interaction might direct formation of a particular complex via multiple cooperative interactions, so that the binding of Gad-SLP-76 might enhance the probability of PLC-γ 1 binding. However, the presence of any particular complex might also depend on local concentrations of proteins, and these factors might depend as well on the state of the cell. To date there has been little analysis of the heterogeneity of LAT-based complexes to address the sort of questions raised above. It should be possible to determine by immunodepletion which molecules can coexist in the same complex. The possibility that multiple, different LAT-based complexes exist is real. Tremendous variety in time and space might be observed. Such different LAT-based complexes might come together so that the sum of complexes would be the critical factor determining progression of a signal for activation. The role of complexes not mediated by LAT will likely receive much attention (99). Signaling events may require the generation of a variety of structures or complexes of complexes to coordinate the various events that occur following TCR engagement.

LAT AND THE CYTOSKELETON To date, most of the studies of adapter molecules involved in lymphocyte signaling have focused on characterizing associated proteins and demonstrating how these interactions regulate classic biochemical signaling cascades. One such pathway involves Grb2, which brings the Ras activator SOS to receptors or to other adapter molecules such as LAT, which have been phosphorylated on tyrosine residues. Activation of Ras leads to subsequent activation in sequence of several serinethreonine kinases, which in turn are responsible for enhanced transcription of a number of genes. (4). Recently there has been increasing attention to another consequence of receptor-mediated signaling, the regulation of the cytoskeleton. Cytoskeletal changes are required for lymphocyte movement, and they accompany and control adhesive interactions that regulate cell-cell interactions (100). These issues have relevance to the early events leading to T cell activation. Of more immediate relevance for this review is the realization that T cell activation involves significant rearrangement of a number of receptors and intracellular molecules over a prolonged period following interaction of the T cell and the ligand-bearing antigen presenting cell (APC). These molecular movements create a supramolecular arrangement of receptors known as the synapse (101, 102). The dynamics of these molecular rearrangements are in part regulated by the cytoskeleton. Another significant and relevant recent breakthrough comes from the basic cell biological study of the cytoskeleton (103–106). Investigators studying several model systems have made a number of conceptual advances in understanding the dynamics of actin polymerization. Many of the molecules involved in this process are either identical to or related to molecules known to interact directly or indirectly with critical lymphocyte adapters during T cell activation. A detailed description of the immune synapse is outside the scope of this manuscript, and the topic has been extensively reviewed. In brief, several groups

1 Feb 2002

10:21

AR

AR152-13.tex

AR152-13.SGM

LaTeX2e(2001/05/10)

P1: GJC

Annu. Rev. Immunol. 2002.20:371-394. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

TCR SIGNALING AND ADAPTER PROTEINS

385

observed, using fluorescence microscopy, that TCR engagement induces a series of molecular rearrangements at the contact zone between a T cell and an antigen presenting cell (APC). Upon T cell contact with an APC bearing peptide-MHC, a central region containing T cell integrin receptors and APC integrin receptor ligands is surrounded by a ring of MHC-peptide complexes. Over minutes this pattern reverses such that the TCR-MHC contacts move to the center. This region is known as the cSMAC (the central supramolecular activation cluster). It is surrounded by the integrin receptors that define the pSMAC or peripheral supramolecular activation cluster (101, 102, 107). Additional studies have located a number of molecules within the context of the SMAC architecture. Thus, for example PKC2 is found in the cSMAC, talin is found in the pSMAC, whereas CD43 is excluded entirely from both cSMAC and pSMAC (101, 108). More recently individual molecules have been shown to migrate in relation to the SMAC over time. The integral membrane tyrosine phosphatase CD45 is initially excluded from the cSMAC and later migrates back to it (109). These supramolecular structures are stable over hours. The importance of actin polymerization to the generation of these structures is confirmed by their disruption after blockade with cytochalasin D. Actin polymerization can be observed using fluorescent phalloidin binding in a microscopic or flow cytometric assay. A ring of polymerized actin can be detected in T cell-APC conjugates or at the interface between T cells and beads coated with stimulatory anti-TCR antibodies. Actin polymerization has recently been observed in live cells using Jurkat T cells stably expressing EGFP-actin (110). In this assay the cells are dropped on to cover slips coated with stimulatory anti-TCR antibodies. Upon contact lamellipodial projections from the cells engage the cover slip and merge into a circumferential ring tightly adherent to the coverslip. This ring spreads outward over 3–5 min as the cell spreads on to the coverslip. The ring is formed of polymerized actin, and rapid polymerization-depolymerization reactions can be observed with fast microscopy systems. This system is also amenable to quantitation. An index of spreading can be calculated by measuring the ratio of actin clustering at the cover slip and the cell body. With this assay actin polymerization appears to be biphasic with an early peak at 3–5 min followed by a prolonged shoulder lasting 15–20 min. The assay can be used to test the effects of various inhibitors on actin polymerization. The role of the LAT molecule can be demonstrated using this assay. Jurkat cells lacking LAT (J.CaM.2 cells) spread very poorly on coverslips coated with stimulatory antibodies, and the little actin polymerization that is seen is short lived. In parallel the assay was also used to evaluate the function of various LAT tyrosine residues and thereby the role of different pathways coupled to LAT. Interestingly no difference appeared in response by these various mutants. The same degree of inhibition of actin polymerization was observed regardless of whether LAT was absent or whether LAT lacked a PLC-γ binding site or lacked all the Grb2 and Gads binding sites. Inhibitors were used to demonstrate a calcium-sensitive component to the regulation of actin polymerization; the inhibition of actin polymerization due to lack of PLC-γ 1 binding may be thus explained. The loss of Grb2 and Gads binding sites may have many consequences leading to problems with actin

1 Feb 2002

10:21

Annu. Rev. Immunol. 2002.20:371-394. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

386

AR

AR152-13.tex

AR152-13.SGM

LaTeX2e(2001/05/10)

P1: GJC

SAMELSON

polymerization. These include loss of optimal PLC-γ 1 binding, as mentioned above, or failure of SLP-76 association, among many possibilities. As described above SLP-76 is a multidomain adapter protein. The central region of SLP-76 contains a proline-rich region that mediates Gads binding. In addition to that interaction and its interaction with PLC-γ 1, SLP-76 also makes contact with a number of proteins that have an impact on the cytoskeleton. The SLP-76 aminoterminus contains tyrosine residues, which after phosphorylation bind Vav and Nck (59, 111, 112). Vav too is a PTK substrate and multidomain protein consisting of PH, SH2, and SH3 binding domains and a Dbl-homology domain required for activation of Rac or cdc42, small G proteins of the Rho family (113–115). Targeted disruption of Vav produces a complex T cell deficit, including a partial block in calcium mobilization and a defect in IL-2 production. Two recent studies demonstrated that T cells from these mice also demonstrate a defect in cytoskeletal function (116, 117). Antibody cross-linking of T cells from these mice failed to produce antigen-receptor caps or patches. A failure of actin polymerization was also demonstrated in a phalloidin-binding assay. The pattern of inhibition was mimicked in these studies by treatment with cytochalasin D. More recently SMAC formation was also shown to be impaired in T cells deficient in Vav (118). The absence of Vav or, in the case of the LAT-deficient cells, the failure of Vav recruitment via Gads and SLP-76 would decrease the amount of activated Rac and cdc42 in the vicinity of the TCR and LAT. Consequences could include inadequate activation of phosphatidylinositol 4-phosphate 5-kinase, which is responsible for generating the PLC-γ 1 substrate phosphatidylinositol 4,5-bisphosphate (PIP2) (119). More importantly for this discussion, lack of activated Rac could result in inadequate WASP activation. WASP is known to bind the Nck adapter molecule, which in turn binds SLP-76 (120). WASP was first identified as the defective protein in patients with WiskottAldrich syndrome (121). T cells from patients with that disease, and murine cells from animals with targeted deletion of the WAS gene, have a phenotype similar to that observed in the Vav −/− animals, showing decreased calcium flux, IL-2 production, and notably, defective actin polymerization (122, 123). An explanation for this phenotype is now clear with an increased understanding of the protein WASP (104, 124). This multidomain protein contains regions capable of binding activated Rac, phospholipids such as PIP2, soluble actin-profilin complexes, and at the C-terminus, the Arp2/3 complex responsible for actin polymerization. In the resting state WASP exists in an autoinhibited state in which the GTPase binding domain interacts with the C-terminal region. Activated G proteins and phospholipids synergistically activate WASP, thus allowing the Arp2/3 complex to mediate actin polymerization. Thus LAT, by recruiting the Gads-SLP-76 complex, may bring together WASP and Vav, the enzyme needed to produce activators of WASP. A third protein bound to SLP-76 and relevant to actin polymerization is SLAP130/Fyb (62a, 62b, 125, 126). This protein was independently isolated as a protein that binds the SLP-76 SH2 domain and as a Fyn-binding protein. A clue to the function of this protein was recently presented (127). These investigators were studying a family of WASP-related proteins and in particular were interested

1 Feb 2002

10:21

AR

AR152-13.tex

AR152-13.SGM

LaTeX2e(2001/05/10)

P1: GJC

Annu. Rev. Immunol. 2002.20:371-394. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

TCR SIGNALING AND ADAPTER PROTEINS

387

in proteins that bound to a particular domain in these proteins, the EVH1 domain. Their studies revealed that SLAP-130/Fyb is one such protein. They then demonstrated that this protein colocalized to polymerized actin in the assay in which Jurkat was activated by anti-CD3 coated beads. In addition they found colocalization with WASP, Arp2/3, Vav, and the EVH1-bearing protein, Evl. Microinjection of peptides capable of blocking the interaction of SLAP130/Fyb with Evl blocked actin remodeling at the bead interface. Similarly, disruption of Arp2/3 localization blocked actin polymerization. The actual function of Evl and other VASP-related proteins in actin polymerization in T cells is not clear. However, these authors demonstrate in this study that SLAP130/Fyb is an additional required component of the actin polymerization machinery in T cells. SLP-76 thus binds the central molecular machinery involved in actin polymerization. Phosphorylation of LAT might bring these SLP-76-bound molecules together with other enzymes such as PLC-γ 1 and PI3K, which may bind LAT directly (91) to the site of TCR engagement and PTK activation. Many of these molecules themselves integrate multiple inputs as described above for WASP and previously for PLC-γ 1. In turn colocalization of these molecules at LAT ensures a highly ordered process of activation. Thus a staggering amount of molecular integration is occurring at multiple levels. In this context it is worth re-asking the question of whether one LAT molecule or one SLP-76 molecule can itself bind all the possible proteins to which it could bind. At the level of the individual protein molecule this seems sterically unlikely. However, mixed populations of molecules and complexes may very well be colocalized at sites of activation. In this view, T cell activation represents the assembly of multiple and varied complexes over time and in particular locations. The challenge for future studies of adapters and T cell activation is to develop approaches and techniques capable of defining these complicated molecular interactions and dynamics.

ACKNOWLEDGMENTS Thanks to Drs. Oreste Acuto, Stephen Bunnel, Claudette Fuller, Jon Houtman, and Connie Sommers for reading the manuscript and making numerous constructive suggestions. Visit the Annual Reviews home page at www.annualreviews.org

LITERATURE CITED 1. Schlessinger J. 2000. Cell signaling by receptor tyrosine kinases. Cell 103:211–25 2. Pawson T. 1995. Protein modules and signalling networks. Nature 373:573–80 3. Pawson T, Scott JD. 1997. Signaling

through scaffold, anchoring, and adaptor proteins. Science 278:2075–80 4. Downward J. 2001. The ins and outs of signalling. Nature 411:759–62 5. Wange RL, Samelson LE. 1996. Complex

1 Feb 2002

10:21

388

6.

7.

Annu. Rev. Immunol. 2002.20:371-394. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

8.

9.

10.

11.

12.

13.

14.

15. 16.

17.

18.

AR

AR152-13.tex

AR152-13.SGM

LaTeX2e(2001/05/10)

P1: GJC

SAMELSON complexes: signaling at the TCR. Immunity 5:197–205 Peterson EJ, Clements JL, Fang N, Koretzky GA. 1998. Adaptor proteins in lymphocyte antigen-receptor signaling. Curr. Opin. Immunol. 10:337–44 Rudd CE. 1999. Adaptors and molecular scaffolds in immune cell signaling. Cell 96:5–8 Samelson LE. 1999. Adaptor proteins and T cell antigen receptor signaling. Prog. Biophys. Mol. Biol. 71:393– 403 Myung PS, Boerthe NJ, Koretzky GA. 2000. Adapter proteins in lymphocyte antigen-receptor signaling. Curr. Opin. Immunol. 12:256–66 Tomlinson MG, Lin J, Weiss A. 2000. Lymphocytes with a complex: adapter proteins in antigen receptor signaling. Immunol. Today 21:584–91 Leo A, Schraven B. 2001. Adapters in lymphocyte signalling. Curr. Opin. Immunol. 13:307–16 Garcia KC, Teyton L, Wilson IA. 1999. Structural basis of T cell recognition. Annu. Rev. Immunol. 17:369–97 Hennecke J, Wiley DC. 2001. T cell receptor-MHC interactions up close. Cell 104:1–4 Ashwell JD, Klausner RD. 1990. Genetic and mutational analysis of the T-cell antigen receptor. Annu. Rev. Immunol. 8:139– 67 Reth M. 1989. Antigen receptor tail clue. Nature 338:383–84 Cambier JC. 1995. New nomenclature for the Reth motif (or ARH1/TAM/ ARAM/YXXL) [letter]. Immunol. Today 16:110 Irving BA, Weiss A. 1991. The cytoplasmic domain of the T cell receptor zeta chain is sufficient to couple to receptorassociated signal transduction pathways. Cell 64:891–901 Letourneur F, Klausner RD. 1991. T-cell and basophil activation through the cytoplasmic tail of T-cell-receptor zeta fam-

19.

20.

21.

22.

23.

24.

25.

26.

ily proteins. Proc. Natl. Acad. Sci. USA 88:8905–9 Straus DB, Weiss A. 1992. Genetic evidence for the involvement of the lck tyrosine kinase in signal transduction through the T cell antigen receptor. Cell 70:585– 93 Gauen LK, Zhu Y, Letourneur F, Hu Q, Bolen JB, Matis LA, Klausner RD, Shaw AS. 1994. Interactions of p59fyn and ZAP-70 with T-cell receptor activation motifs: defining the nature of a signalling motif. Mol. Cell. Biol. 14:3729– 41 Iwashima M, Irving BA, van Oers NSC, Chan AC, Weiss A. 1994. Sequential interactions of the TCR with two distinct cytoplasmic tyrosine kinases. Science 263:1136–39 Wange RL, Malek SN, Desiderio S, Samelson LE. 1993. Tandem SH2 domains of ZAP-70 bind to T cell antigen receptor zeta and CD3 epsilon from activated Jurkat T cells. J. Biol. Chem. 268:19,797–801 Hatada MH, Lu X, Laird ER, Green J, Morgenstern JP, Lou M, Marr CS, Phillips TB, Ram MK, Theriault K, Zoller MJ, Karas JL. 1995. Molecular basis for interaction of the protein tyrosine kinase ZAP-70 with the T-cell receptor. Nature 377:32–38 Wange RL, Guitian R, Isakov N, Watts JD, Aebersold R, Samelson LE. 1995. Activating and inhibitory mutations in adjacent tyrosines in the kinase domain of ZAP-70. J. Biol. Chem. 270:18,730–33 Chan AC, Dalton M, Johnson R, Kong GH, Wang T, Thoma R, Kurosaki T. 1995. Activation of ZAP-70 kinase activity by phosphorylation of tyrosine 493 is required for lymphocyte antigen receptor function. EMBO J. 14:2499–2508 Neumeister EN, Zhu Y, Richard S, Terhorst C, Chan AC, Shaw AS. 1995. Binding of ZAP-70 to phosphorylated Tcell receptor zeta and eta enhances its autophosphorylation and generates specific

1 Feb 2002

10:21

AR

AR152-13.tex

AR152-13.SGM

LaTeX2e(2001/05/10)

P1: GJC

TCR SIGNALING AND ADAPTER PROTEINS

27.

Annu. Rev. Immunol. 2002.20:371-394. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

28.

29.

30.

31.

32.

33.

34.

35.

binding sites for SH2 domain-containing proteins. Mol. Cell. Biol. 15:3171–78 Weiss A, Koretzky G, Schatzman RC, Kadlecek T. 1991. Functional activation of the T-cell antigen receptor induces tyrosine phosphorylation of phospholipase C-gamma 1. Proc. Natl. Acad. Sci. USA 88:5484–88 Secrist JP, Karnitz L, Abraham RT. 1991. T-cell antigen receptor ligation induces tyrosine phosphorylation of phospholipase C-gamma 1. J. Biol. Chem. 266:12,135– 39 June CH, Fletcher MC, Ledbetter JA, Samelson LE. 1990. Increases in tyrosine phosphorylation are detectable before phospholipase C activation after T cell receptor stimulation. J. Immunol. 144:1591–99 Gilliland LK, Schieven GL, Norris NA, Kanner SB, Aruffo A, Ledbetter JA. 1992. Lymphocyte lineage-restricted tyrosine-phosphorylated proteins that bind PLCg SH2 domains. J. Biol. Chem. 267:13610–16 Buday L, Egan SE, Rodriguez Viciana P, Cantrell DA, Downward J. 1994. A complex of Grb2 adaptor protein, Sos exchange factor, and a 36-kDa membranebound tyrosine phosphoprotein is implicated in ras activation in T cells. J. Biol. Chem. 269:9019–23 Sieh M, Bolen JB, Weiss A. 1993. CD45 specifically modulates binding of Lck to a phosphopeptide encompassing the negative regulatory tyrosine of Lck. EMBO J. 12:315–21 Zhang W, Sloan-Lancaster J, Kitchen J, Trible RP, Samelson LE. 1998. LAT: the ZAP-70 tyrosine kinase substrate that links T cell receptor to cellular activation. Cell 92:83–92 Zhang W, Trible RP, Samelson LE. 1998b. LAT palmitoylation: its essential role in membrane microdomain targeting and tyrosine phosphorylation during T cell activation. Immunity 9:239–46 Finco TS, Kadlecek T, Zhang W, Samel-

36.

37.

38.

39.

40.

41.

42.

43.

389

son LE, Weiss A. 1998 LAT is required for TCR-mediated activation of PLCγ1 and the Ras pathway. Immunity 9:617–26 Zhang W, Irvin BJ, Trible RP, Abraham RT, Samelson LE. 1999b. Functional analysis of LAT in TCR-mediated signaling pathways using a LAT-deficient Jurkat cell line. Int. Immunol. 11:943–50 Zhang W, Sommers CL, Burshtyn DN, Stebbins CC, DeJarnette JB, Trible RP, Grinberg A, Tsay HC, Jacobs HM, Kessler CM, Long EO, Love PE, Samelson LE. 1999a. Essential role of LAT in T cell development. Immunity 10:323– 32 Williams BL, Schreiber KL, Zhang W, Wange RL, Samelson LE, Leibson PJ, Abraham RT. 1998. Genetic evidence for differential coupling of Syk family kinases to the T-cell receptor: reconstitution studies in a ZAP-70-deficient Jurkat T-cell line. Mol. Cell. Biol. 18:1388–99 Liu SK, McGlade CJ. 1998. Gads is a novel SH2 and SH3 domain-containing adaptor protein that binds to tyrosinephosphorylated Shc. Oncogene 17:3073– 82 Zhang W, Trible RP, Zhu M, Liu SK, McGlade CJ, Samelson LE. 2000. Association of Grb2, Gads, and phospholipase C-gamma 1 with phosphorylated LAT tyrosine residues. Effect of LAT tyrosine mutations on T cell antigen receptor-mediated signaling. J. Biol. Chem. 275:23,355–61 Lowenstein EJ, Daly RJ, Batzer AG, Li W, Margolis B, Lammers R, Ullrich A, Skolnick EY, Bar-Sagi D, Schlessinger J. 1992. The SH2 and SH3 domaincontaining protein GRB2 links receptor tyrosine kinses to ras signalling. Cell 70: 431–42 Koretzky GA. 1997. The role of Grb2associated proteins in T-cell activation. Immunol. Today 18:401–6 Egan SE, Giddings BW, Brooks MW, Buday L, Sizeland AM, Weinberg RA. 1993. Association of Sos Ras exchange

1 Feb 2002

10:21

390

Annu. Rev. Immunol. 2002.20:371-394. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

44.

45.

46.

47.

48.

49.

50.

51.

52.

AR

AR152-13.tex

AR152-13.SGM

LaTeX2e(2001/05/10)

P1: GJC

SAMELSON protein with Grb2 is implicated in tyrosine kinase signal transduction and transformation. Nature 363:45–51 Nel AE, Gupta S, Lee L, Ledbetter JA, Kanner SB. 1995. Ligation of the T-cell antigen receptor (TCR) induces association of hSos1, ZAP-70, phospholipase Cgamma 1, and other phosphoproteins with Grb2 and the zeta-chain of the TCR. J. Biol. Chem. 270:18,428–36 Dower NA, Stang SL, Bottorff DA, Ebinu JO, Dickie P, Ostergaard HL, Stone JC. 2000. RasGRP is essential for mouse thymocyte differentiation and TCR signaling. Nat. Immunol. 1:317–21 Ebinu JO, Stang SL, Teixeira C, Bottorff DA, Hooton J, Blumberg PM, Barry M, Bleakley RC, Ostergaard HL, Stone JC. 2000. RasGRP links T-cell receptor signaling to Ras. Blood 95:3199–203 Joazeiro CA, Wing SS, Huang H, Leverson JD, Hunter T, Liu YC. 1999. The tyrosine kinase negative regulator c-Cbl as a RING-type, E2-dependent ubiquitinprotein ligase. Science 286:309–12 Ravichandran KS, Lorenz U, Shoelson SE, Burakoff SJ. 1995. Interaction of Shc with Grb2 regulates the Grb2 association with mSOS. Mol. Cell. Biol. 15:593–600 Trub T, Frantz JD, Miyazaki M, Band H, Shoelson SE. 1997. The role of a lymphoid-restricted, Grb2-like SH3SH2-SH3 protein in T cell receptor signaling. J. Biol. Chem. 272:894–902 Liu SK, Fang N, Koretzky GA, McGlade CJ. 1999. The hematopoieticspecific adaptor protein gads functions in T-cell signaling via interactions with the SLP-76 and LAT adaptors. Curr. Biol. 9:67–75 Liu SK, Smith CA, Arnold R, Kiefer F, McGlade CJ. 2000. The adaptor protein Gads (Grb2-related adaptor downstream of Shc) is implicated in coupling hemopoietic progenitor kinase-1 to the activated TCR. J. Immunol. 165:1417–26 Jackman JK, Motto DG, Sun Q, Tanemoto M, Turck CW, Peltz GA, Ko-

53.

54.

55.

56.

57.

58.

59.

60.

retzky GA, Findell PR. 1995. Molecular cloning of SLP-76, a 76-kDa tyrosine phosphoprotein associated with Grb2 in T cells. J. Biol. Chem. 270:7029–32 Motto DG, Ross SE, Wu J, HendricksTaylor LR, Koretzky GA. 1996. Implication of the GRB2-associated phosphoprotein SLP-76 in T cell receptor-mediated interleukin 2 production. J. Exp. Med. 183:1937–43 Musci MA, Motto DG, Ross SE, Fang N, Koretzky GA. 1997. Three domains of SLP-76 are required for its optimal function in a T cell line. J. Immunol. 159:1639– 47 Yablonski D, Kuhne MR, Kadlecek T, Weiss A. 1998. Uncoupling of nonreceptor tyrosine kinases from PLC-gamma1 in an SLP-76-deficient T cell. Science 281:413–16 Pivniouk V, Tsitsikov E, Swinton P, Rathbun G, Alt FW, Geha RS. 1998. Impaired viability and profound block in thymocyte development in mice lacking the adaptor protein SLP-76. Cell 94:229– 38 Clements JL, Yang B, Ross-Barta SE, Eliason SL, Hrstka RF, Williamson RA, Koretzky GA. 1998. Requirement for the leukocyte-specific adapter protein SLP76 for normal T cell development. Science 281:416–19 Boerth NJ, Sadler JJ, Bauer DE, Clements JL, Gheith SM, Koretzky GA. 2000. Recruitment of SLP-76 to the membrane and glycolipid-enriched membrane microdomains replaces the requirement for linker for activation of T cells in T cell receptor signaling. J. Exp. Med. 192:1047–58 Bubeck Wardenburg J, Pappu R, Bu JY, Mayer B, Chernoff J, Straus D, Chan AC. 1998. Regulation of PAK activation and the T cell cytoskeleton by the linker protein SLP-76. Immunity 9:607–16 Su YW, Zhang Y, Schweikert J, Koretzky GA, Reth M, Wienands J. 1999. Interaction of SLP adaptors with the SH2

1 Feb 2002

10:21

AR

AR152-13.tex

AR152-13.SGM

LaTeX2e(2001/05/10)

P1: GJC

TCR SIGNALING AND ADAPTER PROTEINS

61.

Annu. Rev. Immunol. 2002.20:371-394. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

62.

62a.

62b.

63.

64.

65.

66.

67.

68.

domain of Tec family kinases. Eur. J. Immunol. 29:3702–11 Bunnell SC, Diehn M, Yaffe MB, Findell PR, Cantley LC, Berg LJ. 2000. Biochemical interactions integrating Itk with the T cell receptor-initiated signaling cascade. J. Biol. Chem. 275: 2219–30 Liu KQ, Bunnell SC, Gurniak CB, Berg LJ. 1998. T cell receptor-initiated calcium release is uncoupled from capacitative calcium entry in Itk-deficient T cells. J. Exp. Med. 187:1721–27 Griffiths EK, Krawczyk C, Kong YY, Raab M, Hyduk SJ, Bouchard D, Chan VS, Kozieradzki I, Oliveira-dos-Santos AJ, Wakeham A, Ohashi PS, Cybulsky MI, Rudd CE, Penninger JM. 2001. Positive regulation of T cell activation and integrin adhesion by the adapter Fyb/Slap. Science 293:2260–63 Peterson EJ, Woods ML, Dmowski SA, Derimanov G, Jordan MS, Wu JN, Myung PS, Liu QH, Pribila JT, Freedman BD, Shimizu Y, Koretzky GA. 2001. Coupling of the TCR to integrin activation by Slap-130/Fyb. Science 293:2263–65 Brown DA, London E. 1997. Structure of detergent-resistant membrane domains: Does phase separation occur in biological membranes? Biochem. Biophys. Res. Commun. 240:1–7 Simons K, Ikonen E. 1997. Functional rafts in cell membranes. Nature 387:569– 72 Brown DA, London E. 2000. Structure and function of sphingolipid- and cholesterol-rich membrane rafts. J. Biol. Chem. 275:17221–24 von Haller PD, Donohoe S, Aebersold R, Watts JD. 2001. Mass spectrometric characterization of proteins extracted from Jurkat T cell detergent-resistant membrane domains. Eur. J. Mass Spectrom. 7:181–93 Resh MD. 1994. Myristylation and palmitoylation of Src family members: the fats of the matter. Cell 76:411–13 van’t Hof W, Resh MD. 1997. Rapid

69.

70.

71.

72.

73.

74.

75.

76.

77.

78.

391

plasma membrane anchoring of newly synthesized p59fyn: selective requirement for NH2-terminal myristoylation and palmitoylation at cysteine-3. J. Cell. Biol. 136:1023–35 Kabouridis PS, Magee AI, Ley SC. 1997. S-acylation of LCK protein tyrosine kinase is essential for its signalling function in T lymphocytes. EMBO J. 16:4983–98 Field KA, Holowka D, Baird B. 1995. Fc epsilon RI-mediated recruitment of p53/56lyn to detergent-resistant membrane domains accompanies cellular signaling. Proc. Natl. Acad. Sci. USA 92: 9201–5 Field KA, Holowka D, Baird B. 1997. Compartmentalized activation of the high affinity immunoglobulin E receptor within membrane domains. J. Biol. Chem. 272:4276–80 Xavier R, Brennan T, Li Q, McCormack C, Seed B. 1998. Membrane compartmentation is required for efficient T cell activation. Immunity 8:723–32 Montixi C, Langlet C, Bernard AM, Thimonier J, Dubois C, Wurbel MA, Chauvin JP, Pierres M, He HT. 1998. Engagement of T cell receptor triggers its recruitment to low-density detergentinsoluble membrane domains. EMBO J. 17:5334–48 Janes PW, Ley SC, Magee AI. 1999. Aggregation of lipid rafts accompanies signaling via the T cell antigen receptor. J. Cell. Biol. 147:447–61 Lin J, Weiss A, Finco TS. 1999. Localization of LAT in glycolipid-enriched microdomains is required for T cell activation. J. Biol. Chem. 274:28,861–64 Reich Z, Boniface JJ, Lyons DS, Borochov N, Wachtel EJ, Davis MM. 1997. Ligand-specific oligomerization of T-cell receptor molecules. Nature 387:617–20 Duplay P, Thome M, Herve F, Acuto O. 1994. p56lck interacts via its src homology 2 domain with the ZAP-70 kinase. J. Exp. Med. 179:1163–72 Pelosi M, Di Bartolo V, Mounier V, Mege

1 Feb 2002

10:21

392

Annu. Rev. Immunol. 2002.20:371-394. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

79.

80.

81.

82.

83.

84.

85.

86.

AR

AR152-13.tex

AR152-13.SGM

LaTeX2e(2001/05/10)

P1: GJC

SAMELSON D, Pascussi JM, Dufour E, Blondel A, Acuto O. 1999. Tyrosine 319 in the interdomain B of ZAP-70 is a binding site for the src homology 2 domain of Lck. J. Biol. Chem. 274:14,229–37 Bosselut R, Zhang W, Ashe JM, Kopacz JL, Samelson LE, Singer A. 1999. Association of the adaptor molecule LAT with CD4 and CD8 coreceptors identifies a new coreceptor function in T cell receptor signal transduction. J. Exp. Med. 190:1517– 26 Li S, Satoh T, Korngold R, Huang Z. 1998. CD4 dimerization and oligomerization: implications for T-cell function and structure-based drug design. Immunol. Today 19:455–62 Williams BL, Irvin BJ, Sutor SL, Chini CC, Yacyshyn E, Bubeck Wardenburg J, Dalton M, Chan AC, Abraham RT. 1999. Phosphorylation of tyr319 in ZAP70 is required for T-cell antigen receptordependent phospholipase C-γ1 and ras activation. Embo J. 18:1832–44 Deckert M, Tartare-Deckert S, Hernandez J, Rottapel R, Altman A. 1998. Adaptor function for the Syk kinasesinteracting protein 3BP2 in IL-2 gene activation. Immunity 9:595–605 Lindholm CK, Gylfe E, Zhang W, Samelson LE, Welsh M. 1999. Requirement of the Src homology 2 domain protein Shb for T cell receptor-dependent activation of the interleukin-2 gene nuclear factor for activation of T cells element in Jurkat T cells. J. Biol. Chem. 274:28,050– 57 Harder T, Kuhn M. 2000. Selective accumulation of raft-associated membrane protein LAT in T cell receptor signaling assemblies. J. Cell. Biol. 151:199–208 Varma R, Mayor S. 1998. GPI-anchored proteins are organized in submicron domains at the cell surface. Nature 394:798– 801 Kenworthy AK, Petranova N, Edidin M. 2000. High-resolution FRET microscopy of cholera toxin B-subunit and GPI-an-

87.

88.

89.

90.

91.

92.

93.

94.

95.

chored proteins in cell plasma membranes. Mol. Biol. Cell. 11:1645–55 Nishibe S, Wahl MI, HernandezSotomayor SM, Tonks NK, Rhee SG, Carpenter G. 1990. Increase of the catalytic activity of phospholipase C-gamma 1 by tyrosine phosphorylation. Science 250:1253–56 Rhee SG, Bae YS. 1997. Regulation of phosphoinositide-specific phospholipase C isoenzymes. J. Biol. Chem. 272:15,045–48 Weber JR, Bell GM, Han MY, Pawson T, Imboden JB. 1992. Association of the tyrosine kinase LCK with phospholipase C-γ1 after stimulation of the T cell antigen receptor. J. Exp. Med. 176:373–79 Stoica B, DeBell KE, Graham L, Rellahan BL, Alava MA, Laborda J, Bonvini E. 1998. The amino-terminal Src homology 2 domain of phospholipase C-γ1 is essential for TCR-induced tyrosine phosphorylation of phospholipase C-γ1. J. Immunol. 160:1059–66 Paz PE, Wang S, Clarke H, Lu X, Stokoe D, Abo A. 2001. Mapping the Zap-70 phosphorylation sites on LAT (linker for activation of T cells) required for recruitment and activation of signalling proteins in T cells. Biochem. J. 356: 461–71 Lin J, Weiss A. 2001. Identification of the minimal tyrosine residues required for LAT function. J. Biol. Chem. 276:29,588– 95 Yablonski D, Kadlecek T, Weiss A. 2001. Identification of a phospholipase C-gamma1 (PLC-gamma1) SH3 domainbinding site in SLP-76 required for Tcell receptor-mediated activation of PLCgamma1 and NFAT. Mol. Cell. Biol. 21:4208–18 Schaeffer EM, Debnath J, Yap G, McVicar D, Liao XC, Littman DR, Sher A, Varmus HE, Lenardo MJ, Schwartzberg PL. 1999. Requirement for Tec kinases Rlk and Itk in T cell receptor signaling and immunity. Science 284:638–41 Sommers CL, Rabin RL, Grinberg A,

1 Feb 2002

10:21

AR

AR152-13.tex

AR152-13.SGM

LaTeX2e(2001/05/10)

P1: GJC

TCR SIGNALING AND ADAPTER PROTEINS

Annu. Rev. Immunol. 2002.20:371-394. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

96.

97.

98.

99.

100.

101.

102.

103.

104.

105.

106.

Tsay HC, Farber J, Love PE. 1999. A role for the Tec family tyrosine kinase Txk in T cell activation and thymocyte selection. J. Exp. Med. 190:1427–38 Shan X, Wange RL. 1999. Itk/Emt/Tsk activation in response to CD3 crosslinking in Jurkat T cells requires ZAP-70 and Lat and is independent of membrane recruitment. J. Biol. Chem. 274:29,323– 30 Ladbury JE, Arold S. 2000. Searching for specificity in SH domains. Chem. Biol. 7:R3–8 Liou J, Kiefer F, Dang A, Hashimoto A, Cobb MH, Kurosaki T, Weiss A. 2000. HPK1 is activated by lymphocyte antigen receptors and negatively regulates AP-1. Immunity 12:399–408 Ku GM, Yablonski D, Manser E, Lim L, Weiss A. 2001. A PAK1-PIX-PKL complex is activated by the T-cell receptor independent of Nck, Slp-76 and LAT. EMBO J. 20:457–65 Dustin ML, Cooper JA. 2000. The immunological synapse and the actin cytoskeleton: molecular hardware for T cell signaling. Nat. Immunol. 1:23–29 Monks CR, Freiberg BA, Kupfer H, Sciaky N, Kupfer A. 1998. Three-dimensional segregation of supramolecular activation clusters in T cells. Nature 395: 82–86 Grakoui A, Bromley SK, Sumen C, Davis MM, Shaw AS, Allen PM, Dustin ML. 1999. The immunological synapse: a molecular machine controlling T cell activation. Science 285:221–27 Machesky LM, Insall RH. 1999. Signaling to actin dynamics. J. Cell. Biol. 146:267–72 Higgs HN, Pollard TD. 1999. Regulation of actin polymerization by Arp2/3 complex and WASp/Scar proteins. J. Biol. Chem. 274:32,531–34 Borisy GG, Svitkina TM. 2000. Actin machinery: pushing the envelope. Curr. Opin. Cell Biol. 12:104–12 Pantaloni D, Le Clainche C, Carlier MF.

107.

108.

109.

110.

111.

112.

113.

114.

115.

116.

393

2001. Mechanism of actin-based motility. Science 292:1502–6 Dustin ML, Chan AC. 2000. Signaling takes shape in the immune system. Cell 103:283–94 Sperling AI, Sedy JR, Manjunath N, Kupfer A, Ardman B, Burkhardt JK. 1998. TCR signaling induces selective exclusion of CD43 from the T cell-antigenpresenting cell contact site. J. Immunol. 161:6459–62 Johnson KG, Bromley SK, Dustin ML, Thomas ML. 2000. A supramolecular basis for CD45 tyrosine phosphatase regulation in sustained T cell activation. Proc. Natl. Acad. Sci. USA 97:10,138–43 Bunnell SC, Kapoor V, Trible RP, Zhang W, Samelson LE. 2001. Dynamic actin polymerization drives T cell receptorinduced spreading: a role for the signal transduction adaptor LAT. Immunity 14:315–29 Tuosto L, Michel F, Acuto O. 1996. p95vav associates with tyrosine-phosphorylated SLP-76 in antigen-stimulated T cells. J. Exp. Med. 184:1161–66 Wunderlich L, Farago A, Downward J, Buday L. 1999. Association of Nck with tyrosine-phosphorylated SLP-76 in activated T lymphocytes. Eur. J. Immunol. 29:1068–75 Bustelo XR, Ledbetter JA, Barbacid M. 1992. Product of vav proto-oncogene defines a new class of tyrosine protein kinase substrates. Nature 356:68–71 Margolis B, Hu P, Katzav S, Li W, Oliver JM, Ullrich A, Weiss A, Schlessinger J. 1992. Tyrosine phosphorylation of vav proto-oncogene product containing SH2 domain and transcription factor motifs. Nature 356:71–74 Crespo P, Schuebel KE, Ostrom AA, Gutkind JS, Bustelo XR. 1997. Phosphotyrosine-dependent activation of Rac-1 GDP/GTP exchange by the vav protooncogene product. Nature 385:169–72 Fischer KD, Kong YY, Nishina H, Tedford K, Marengere LE, Kozieradzki I,

1 Feb 2002

10:21

394

Annu. Rev. Immunol. 2002.20:371-394. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

117.

118.

119.

120.

121.

122.

AR

AR152-13.tex

AR152-13.SGM

LaTeX2e(2001/05/10)

P1: GJC

SAMELSON Sasaki T, Starr M, Chan G, Gardener S, Nghiem MP, Bouchard D, Barbacid M, Bernstein A, Penninger JM. 1998. Vav is a regulator of cytoskeletal reorganization mediated by the T-cell receptor. Curr. Biol. 8:554–62 Holsinger LJ, Graef IA, Swat W, Chi T, Bautista DM, Davidson L, Lewis RS, Alt FW, Crabtree GR. 1998. Defects in actin-cap formation in Vav-deficient mice implicate an actin requirement for lymphocyte signal transduction. Curr. Biol. 8:563–72 Wulfing C, Bauch A, Crabtree GR, Davis MM. 2000. The vav exchange factor is an essential regulator in actin-dependent receptor translocation to the lymphocyteantigen-presenting cell interface. Proc. Natl. Acad. Sci. USA 97:10,150–55 O’Rourke LM, Tooze R, Turner M, Sandoval DM, Carter RH, Tybulewicz VL, Fearon DT. 1998. CD19 as a membraneanchored adaptor protein of B lymphocytes: costimulation of lipid and protein kinases by recruitment of Vav. Immunity 8:635–45 Rivero-Lezcano OM, Marcilla A, Sameshima JH, Robbins KC. 1995. WiskottAldrich syndrome protein physically associates with Nck through Src homology 3 domains. Mol. Cell. Biol. 15:5725–31 Snapper SB, Rosen FS. 1999. The Wiskott-Aldrich syndrome protein (WASP): roles in signaling and cytoskeletal organization. Annu. Rev. Immunol. 17:905–29 Snapper SB, Rosen FS, Mizoguchi E, Cohen P, Khan W, Liu CH, Hagemann TL, Kwan SP, Ferrini R, Davidson L, Bhan AK, Alt FW. 1998. Wiskott-Aldrich

123.

124.

125.

126.

127.

syndrome protein-deficient mice reveal a role for WASP in T but not B cell activation. Immunity 9:81–91 Zhang J, Shehabeldin A, da Cruz LA, Butler J, Somani AK, McGavin M, Kozieradzki I, dos Santos AO, Nagy A, Grinstein S, Penninger JM, Siminovitch KA. 1999. Antigen receptor-induced activation and cytoskeletal rearrangement are impaired in Wiskott-Aldrich syndrome protein-deficient lymphocytes. J. Exp. Med. 190:1329–42 Mullins RD. 2000. How WASP-family proteins and the Arp2/3 complex convert intracellular signals into cytoskeletal structures. Curr. Opin. Cell Biol. 12:91– 96 Musci MA, Hendricks-Taylor LR, Motto DG, Paskind M, Kamens J, Turck CW, Koretzky GA. 1997. Molecular cloning of SLAP-130, an SLP-76-associated substrate of the T cell antigen receptorstimulated protein tyrosine kinases. J. Biol. Chem. 272:11674–77 da Silva AJ, Li Z, de Vera C, Canto E, Findell P, Rudd CE. 1997. Cloning of a novel T-cell protein FYB that binds FYN and SH2-domain-containing leukocyte protein 76 and modulates interleukin 2 production. Proc. Natl. Acad. Sci. USA 94:7493–98 Krause M, Sechi AS, Konradt M, Monner D, Gertler FB, Wehland J. 2000. Fynbinding protein (Fyb)/SLP-76-associated protein (SLAP), Ena/vasodilator-stimulated phosphoprotein (VASP) proteins and the Arp2/3 complex link T cell receptor (TCR) signaling to the actin cytoskeleton. J. Cell. Biol. 149:181–94

8:19 AR AR152-13-COLOR.tex AR152-13-COLOR.SGM

Figure 1 A selection of signaling proteins found in T lymphocytes depicted to highlight their modular structures. SH2, SH3 and PH domains are in red, blue and green respectively. A Tec homology domain is in pink, and a transmembrane domain is in light green. Sites of tyrosine phosphorylation are indicated with Y and proline-rich sites are indicated Pro. Domains with enzymatic function are in yellow.

Annu. Rev. Immunol. 2002.20:371-394. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

6 Feb 2002 LaTeX2e(2002/01/18) P1: GDL

8:19 AR AR152-13-COLOR.tex AR152-13-COLOR.SGM

Figure 2 Signaling complexes at the T cell antigen receptor and at the LAT molecule. The TCR is depicted with its associated protein tyrosine kinases. Tyrosine phosphates are indicated with a black ball and the modular components of other linkers and enzymes are color coded with SH2 domains in red, SH3 in light blue, PH in green, proline-rich regions in dark blue, and enzymatic domains in yellow.

Annu. Rev. Immunol. 2002.20:371-394. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

6 Feb 2002 LaTeX2e(2002/01/18) P1: GDL

P1: FRK

February 1, 2002

12:47

Annual Reviews

AR152-FM

Annual Review of Immunology Volume 20, 2002

Annu. Rev. Immunol. 2002.20:371-394. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

CONTENTS FRONTISPIECE, Charles A. Janeway, Jr. A TRIP THROUGH MY LIFE WITH AN IMMUNOLOGICAL THEME, Charles A. Janeway, Jr.

xiv 1

THE B7 FAMILY OF LIGANDS AND ITS RECEPTORS: NEW PATHWAYS FOR COSTIMULATION AND INHIBITION OF IMMUNE RESPONSES, Beatriz M. Carreno and Mary Collins

29

MAP KINASES IN THE IMMUNE RESPONSE, Chen Dong, Roger J. Davis, and Richard A. Flavell

PROSPECTS FOR VACCINE PROTECTION AGAINST HIV-1 INFECTION AND AIDS, Norman L. Letvin, Dan H. Barouch, and David C. Montefiori T CELL RESPONSE IN EXPERIMENTAL AUTOIMMUNE ENCEPHALOMYELITIS (EAE): ROLE OF SELF AND CROSS-REACTIVE ANTIGENS IN SHAPING, TUNING, AND REGULATING THE AUTOPATHOGENIC T CELL REPERTOIRE, Vijay K. Kuchroo, Ana C. Anderson, Hanspeter Waldner, Markus Munder, Estelle Bettelli, and Lindsay B. Nicholson

55 73

101

NEUROENDOCRINE REGULATION OF IMMUNITY, Jeanette I. Webster, Leonardo Tonelli, and Esther M. Sternberg

125

MOLECULAR MECHANISM OF CLASS SWITCH RECOMBINATION: LINKAGE WITH SOMATIC HYPERMUTATION, Tasuku Honjo, Kazuo Kinoshita, and Masamichi Muramatsu

165

INNATE IMMUNE RECOGNITION, Charles A. Janeway, Jr. and Ruslan Medzhitov

KIR: DIVERSE, RAPIDLY EVOLVING RECEPTORS OF INNATE AND ADAPTIVE IMMUNITY, Carlos Vilches and Peter Parham ORIGINS AND FUNCTIONS OF B-1 CELLS WITH NOTES ON THE ROLE OF CD5, Robert Berland and Henry H. Wortis E PROTEIN FUNCTION IN LYMPHOCYTE DEVELOPMENT, Melanie W. Quong, William J. Romanow, and Cornelis Murre

197 217 253 301

LYMPHOCYTE-MEDIATED CYTOTOXICITY, John H. Russell and Timothy J. Ley

SIGNAL TRANSDUCTION MEDIATED BY THE T CELL ANTIGEN x

323

P1: FRK

Annu. Rev. Immunol. 2002.20:371-394. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

February 1, 2002

12:47

Annual Reviews

AR152-FM

CONTENTS

xi

RECEPTOR: THE ROLE OF ADAPTER PROTEINS, Lawrence E. Samelson INTERACTION OF HEAT SHOCK PROTEINS WITH PEPTIDES AND ANTIGEN PRESENTING CELLS: CHAPERONING OF THE INNATE AND ADAPTIVE IMMUNE RESPONSES, Pramod Srivastava CHROMATIN STRUCTURE AND GENE REGULATION IN THE IMMUNE SYSTEM, Stephen T. Smale and Amanda G. Fisher PRODUCING NATURE’S GENE-CHIPS: THE GENERATION OF PEPTIDES FOR DISPLAY BY MHC CLASS I MOLECULES, Nilabh Shastri, Susan

371

Schwab, and Thomas Serwold

395 427

463

THE IMMUNOLOGY OF MUCOSAL MODELS OF INFLAMMATION, Warren Strober, Ivan J. Fuss, and Richard S. Blumberg T CELL MEMORY, Jonathan Sprent and Charles D. Surh

GENETIC DISSECTION OF IMMUNITY TO MYCOBACTERIA: THE HUMAN MODEL, Jean-Laurent Casanova and Laurent Abel ANTIGEN PRESENTATION AND T CELL STIMULATION BY DENDRITIC CELLS, Pierre Guermonprez, Jenny Valladeau, Laurence Zitvogel, Clotilde Th´ery, and Sebastian Amigorena

495 551 581

621

NEGATIVE REGULATION OF IMMUNORECEPTOR SIGNALING, Andr´e Veillette, Sylvain Latour, and Dominique Davidson

669

CPG MOTIFS IN BACTERIAL DNA AND THEIR IMMUNE EFFECTS, Arthur M. Krieg

PROTEIN KINASE Cθ

709 IN

T CELL ACTIVATION, Noah Isakov and Amnon

Altman

RANK-L AND RANK: T CELLS, BONE LOSS, AND MAMMALIAN EVOLUTION, Lars E. Theill, William J. Boyle, and Josef M. Penninger PHAGOCYTOSIS OF MICROBES: COMPLEXITY IN ACTION, David M. Underhill and Adrian Ozinsky

761 795 825

STRUCTURE AND FUNCTION OF NATURAL KILLER CELL RECEPTORS: MULTIPLE MOLECULAR SOLUTIONS TO SELF, NONSELF DISCRIMINATION, Kannan Natarajan, Nazzareno Dimasi, Jian Wang, Roy A. Mariuzza, and David H. Margulies

853

INDEXES Subject Index Cumulative Index of Contributing Authors, Volumes 1–20 Cumulative Index of Chapter Titles, Volumes 1–20

ERRATA An online log of corrections to Annual Review of Immunology chapters (if any, 1997 to the present) may be found at http://immunol.annualreviews.org/

887 915 925

8 Feb 2002

9:57

AR

AR152-14.tex

AR152-14.SGM

LaTeX2e(2001/05/10) P1: GKD 10.1146/annurev.immunol.20.100301.064801

Annu. Rev. Immunol. 2002. 20:395–425 DOI: 10.1146/annurev.immunol.20.100301.064801 c 2002 by Annual Reviews. All rights reserved Copyright °

INTERACTION OF HEAT SHOCK PROTEINS WITH PEPTIDES AND ANTIGEN PRESENTING CELLS: Annu. Rev. Immunol. 2002.20:395-425. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

Chaperoning of the Innate and Adaptive Immune Responses Pramod Srivastava Center for Immunotherapy of Cancer and Infectious Diseases, University of Connecticut School of Medicine, Farmington, Connecticut 06030-1601; e-mail: [email protected]

Key Words dendritic cells, cross-priming, indirect presentation, cancer, infectious diseases ■ Abstract Heat shock proteins are abundant soluble intracellular proteins, present in all cells. Members of the heat shock protein family bind peptides including antigenic peptides generated within cells. Heat shock proteins also interact with antigen presenting cells through CD91 and other receptors, eliciting a cascade of events including re-presentation of heat shock protein-chaperoned peptides by MHC, translocation of NFκB into the nuclei and maturation of dendritic cells. These consequences point to a key role of heat shock proteins in fundamental immunological phenomena such as activation of antigen presenting cells, indirect presentation (or cross-priming), and chaperoning of peptides during antigen presentation. Heat shock proteins appear to have been involved in innate immune responses since the emergence of phagocytes in early multicellular organisms and to have been commandeered for adaptive immune responses with the advent of specificity. These properties of heat shock proteins also allow them to be used for immunotherapy of cancers and infections in novel ways.

THE MHC AND THE HEAT SHOCK PROTEINS, A COMMON PEDIGREE Transplantation of tissues and tumors among mice led to the identification of fundamental immunological roles for two major groups of molecules, the MHC and the heat shock proteins (HSPs). At first sight, the HSPs and the MHC proteins appear quite dissimilar. The MHC proteins are of very recent evolutionary vintage, while the HSPs appeared at the very dawn of life. The MHC are among the most polymorphic (poly-allelic) proteins, while the HSPs are typically monoallelic. The MHC are cell surface proteins, whereas the HSPs are essentially intracellular. The MHC are expressed at modest levels while the HSPs are embarrassingly abundant. However, 0732-0582/02/0407-0395$14.00

395

4 Feb 2002

11:36

Annu. Rev. Immunol. 2002.20:395-425. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

396

AR

AR152-14.tex

AR152-14.SGM

LaTeX2e(2001/05/10)

P1: GKD

SRIVASTAVA

these significant differences hide powerful similarities and convergence of functions. In fact, the HSPs have been among the key players in host defense for billenia and appear to have been laying the groundwork for many of the latter day functions of the MHC molecules. I aim to tell here the story of the primordial functions of the HSPs in innate immunity and to describe the many paths where the functions of the MHC proteins and the HSPs converge to create the symphony of adaptive immunity. While the MHC molecules are familiar to immunologists, a brief introduction to HSPs may not be out of place (1). Approximately 40 years ago, somebody inadvertently turned up the temperature of an incubator full of fruit flies, and the salivary gland chromosomes of the fruit flies, thus heat-shocked, showed the characteristic puffs indicative of transcriptional activity at discrete loci (2). These loci came to be known to encode HSPs, which were gradually identified in all species tested. They are expressed in all cells in all forms of life and in a variety of intracellular locations: in the cytosol of prokaryotes and in the cytosol, nuclei, endoplasmic reticulum, mitochondria, and chloroplasts of eukaryotes. In addition to their ubiquity, the HSPs constitute the single most abundant group of proteins inside cells. They are expressed in vast quantities under normal non–heat shocked conditions, and their expression can be powerfully induced to much higher levels as a result of heat shock or other forms of stress, including exposure to toxins, oxidative stress, glucose deprivation, etc. Approximately ten families of HSPs are known, and each family consists of anywhere from 1 to 5 closely related proteins. There is little or no obvious homology among the individual HSP families even as members within a family are closely related. All HSP families are represented in all organisms although individual members may show variety in distribution. Since their discovery, an increasing array of functions such as folding and unfolding of proteins (3), degradation of proteins (4), assembly of multi-subunit complexes (5), thermotolerance (6), buffering of expression of mutations (7), and others have been attributed to HSPs. In addition, they have become absorbing models for the study of transcriptional regulation (8), stress response (6), and evolution (9). In order to tell the story of the common experimental pedigree of the MHC and the HSPs, I begin with a quote from a recent article by George Klein that summarizes the beginnings of the connection between the MHC molecules and tumor immunity (10): During the first part of the 20th century, cancer researchers spoke about transplantable and non-transplantable tumors. The mice and rats were not inbred and transplantability meant therefore transgression of histocompatibility barriers, but most researchers were unaware of this . . . tumor immunology was an optimistic field due to this artifact. In spite of this clear, definitive evidence that was available already in the early 50s, the artifactual, allograft-based “tumor immunology” continued to flourish during at least one more decade. Meanwhile Ludwik Gross performed some not too well controlled experiments suggesting that chemically induced mouse sarcomas could be immunogenic in syngeneic mice. Subsequently Prehn and Main confirmed this in critically controlled experiments . . . Their data also indicated that the chemically

4 Feb 2002

11:36

AR

AR152-14.tex

AR152-14.SGM

LaTeX2e(2001/05/10)

P1: GKD

ROLE OF HSPs IN IMMUNE RESPONSE

397

induced tumors did not cross-react with each other . . . We suspected that even the experiments of Prehn and Main may have been flawed. Obviously, the ultimate evidence had to be based on experiments with the primary, autochthonous tumor host. We did these rather laborious experiments and published them in 1960 in Cancer Research. Yes, it was all true.

Annu. Rev. Immunol. 2002.20:395-425. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

Figure 1 shows the general outline of the experiments of Klein et al. (11). They demonstrated the extraordinary phenomenon that one could immunize against

Figure 1 A cartoon showing the design of the experiments of Klein et al. (11) that established formally that methylcholanthrene-induced tumors are immunogenic in the primary and in the syngeneic hosts, and that immunity is individually tumor-specific.

4 Feb 2002

11:36

Annu. Rev. Immunol. 2002.20:395-425. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

398

AR

AR152-14.tex

AR152-14.SGM

LaTeX2e(2001/05/10)

P1: GKD

SRIVASTAVA

syngeneic tumors in the same manner as one could against smallpox and polio viruses. The extraordinariness derived from the fact that in contrast to viruses, the tumors were of self origin and yet were immunogenic. Inherent in this observation was the prediction that the tumors express tumor-specific antigens. Further, as the individual tumors were not cross-reactive, the tumor-specific antigens were individually rather than commonly tumor-specific. Speaking with respect to the identity of these antigens, referred to at the time as tumor-specific transplantation antigens (TSTAs), Klein ends his article with the accurate remark “The TSTAs of the chemically induced tumors are still a mystery.” I began to look for the cancer-specific antigens by their ability to elicit protective immunity to cancer challenges, i.e., by the very assay that pointed to their existence (see 12). This approach typically involved fractionation of cancer homogenates into various protein components by conventional chromatographic methods. The fractions thus obtained were used to immunize animals that were then challenged with live cancer cells. The fractions that elicited protection against the cancer were then re-fractionated and the cancer rejection assay repeated until apparently homogeneous preparations were obtained. This approach, with variations, led to identification of cancer-rejection molecules from cancers of diverse histological origins, induced in mice and rats of different haplotypes by chemicals or UV-radiation, or they were of spontaneous origin (Table 1). The cancers ranged in immunogenicity from the nonimmunogenic (e.g., the Lewis lung carcinoma) to the highly immunogenic regressor cancers induced by UV-radiation. Surprisingly, all the well-characterized molecules identified by this method, by us and then by others, turned out to be HSPs of the hsp90, hsp70, calreticulin, or the grp170 family (Table 1). The phenomenon of graft and tumor rejection among histo-incompatble mice played a key role in the discovery of histocompatibility and the molecules that mediate it, i.e., the MHC molecules. The same phenomenon in histocompatible (syngeneic) mice led to the discovery of the molecules that mediate such rejection, i.e., the HSPs.

THE STRANGE IMMUNOGENETICITY OF VERY COMMON MOLECULES: DISCOVERY OF HSP-ASSOCIATED PEPTIDES Consistent with the experiments with intact tumors, the HSPs purified from a given cancer were observed to elicit protective immunity specific to that particular cancer. HSPs derived from normal tissues did not elicit protective immunity to any cancers tested (13). The observed specificity of immunogenicity of cancer-derived HSPs suggested that HSPs ought to harbor somatic polymorphisms, such that HSPs would differ between cancers and normal tissues and from one cancer to another. However, extensive sequencing studies of HSP cDNAs of cancers and normal tissues did not support that idea (14). What then was the basis of the specificity of immunogenicity of these very common HSP molecules? The first

4 Feb 2002

11:36

AR

AR152-14.tex

AR152-14.SGM

LaTeX2e(2001/05/10)

P1: GKD

ROLE OF HSPs IN IMMUNE RESPONSE

399

Annu. Rev. Immunol. 2002.20:395-425. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

TABLE 1 Representative studies that demonstrate the tumor-specific immunoprotective activity of tumor-derived HSP-peptide complexes, or infectious disease-specific immunoprotective activity of HSP-peptide complexes derived from infected cells. Immunogenicity of tumors is graded as −,+,++ or +++ by subjective criteria. Asterisks refer to models of therapy of pre-existing cancers, as opposed to models of prophylaxis. Cancer or infectious agent

Induced by

Immunogenicity

Host

Molecule

Ref.

Zajdela hepatoma

Chemical

++

Rat

gp96

39

Meth A fibrosarcoma

Chemical

++

BALB/c mice

gp96 hsp90 hsp70 hsp110 grp170

50, 68*, 94* 98 18 99 99

CMS5

Chemical

+

BALB/c mice

gp96

50

CMS13

Chemical

++

BALB/c mice

gp96

100

Lewis lung ca.

Spontaneous

−

C57BL/6 mice

gp96 hsp70

68*, 94* 68*

B16 melanoma

Spontaneous

−

C57BL/6 mice

gp96

68*, 101*

CT26 colon ca.

Chemical

++

BALB/c mice

gp96

68*

BALB/c mice

hsp110 grp170

99* 99

Colon 26 Ca. UV6138

UV

+++

C3H mice

gp96

102

UV6139SJ

UV

++

C3H mice

gp96

68*, 102

Dunning G prostate ca.

+

Rat

gp96

103*

A20 B cell lymphoma

+

BALB/c mice

gp96 hsp70 hsp90 Calreticulin

104 104 104 104

+

Xenopus

Gp96 Hsp70

90 90

M. tuberculosis

BALB/c mice

Gp96

22

Listeria

BALB/c mice

Gp96

22

LCMV

BALB/c mice

Hsp70

105

L15/0 lymphoma

Spontaneous

obvious answer lay in the possibility that the homogeneous HSP preparations were not so homogeneous after all, but contained unexamined contaminants that were responsible for the immunogenicity. This possibility, as sensible as it was depressing, did not turn out to be true: The immunogenic HSP preparations were certifiably free of other protein contaminants as determined by all structural criteria tested, and the immunogenicity did not derive from associated carbohydrates, lipids, or nuclei acids (P. Srivastava, unpublished observations). The possibility was then envisaged that low molecular weight substances, not detectable by polyacrylamide gel electrophoresis, are associated with HSPs and are responsible for the specificity of immunogenicity of HSP preparations (15, 16). This idea was tested and

4 Feb 2002

11:36

Annu. Rev. Immunol. 2002.20:395-425. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

400

AR

AR152-14.tex

AR152-14.SGM

LaTeX2e(2001/05/10)

P1: GKD

SRIVASTAVA

derived a modicum of support when a large collection of peptides could be shown to elute from a homogeneous gp96 preparation as it was treated with trifluoroacetic acid (17). Strong support for the idea came when treatment of an immunogenic (tumor-protective) hsp70 preparation with ATP had two consequences: It resulted in elution of a wide array of peptide peaks from the hsp70 polypeptide, leaving the polypeptide intact, and it rendered the hsp70 preparation nonimmunogenic, i.e., ineffective in immunizing against cancer cells, even though the hsp70 polypeptide was present in equivalent amounts in untreated and ATP-treated hsp70 preparations (18). This was the first demonstration that hsp70, as isolated from tumors, was associated with peptides, that dissociation of peptides from hsp70 resulted in abrogation of the immunogenicity, and that the hsp70 polypeptide was not immunogenic in and of itself. It was shown subsequently that the hsp70 and gp96 are associated with peptides in vivo, and the observed association of hsp70 and gp96 with peptides is not the result of an artifact occurring after cell lysis and during purification of the HSPs (19). Considerable immunological and structural evidence now supports the notion that certain HSP molecules (gp96, hsp90, hsp70, calreticulin, hsp110, and grp170) are peptide-binding proteins and are associated with antigenic epitopes (Tables 1, 2).

The Immunological Evidence The immunological evidence for association of HSPs and antigenic peptides has continued to accumulate at an impressive pace. The large number of studies that TABLE 2 Selected structural or immunological studies that have shown that specific, defined antigenic peptides are associated with HSPs Epitope/antigen

MHC I

Ref.

TUMOR ANTIGENS PRL1a mouse leukemia Human melanoma MART-1 Human melanoma tyrosinase Human melanoma gp100

Ld A2 A2 A2

28 26 26 26

Kb d Kb Db, Kb

23, 27 25 20 106 30

MODEL ANTIGENS β galactosidase Ovalbumin

Ld Kb

24 29

NORMAL CELLULAR ANTIGENS Minor H

Kd, Kb

24

VIRAL ANTIGENS Vescicular stomatitis Herpes simplex-2 Influenza SV40 Hepatitis B ag

4 Feb 2002

11:36

AR

AR152-14.tex

AR152-14.SGM

LaTeX2e(2001/05/10)

P1: GKD

Annu. Rev. Immunol. 2002.20:395-425. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

ROLE OF HSPs IN IMMUNE RESPONSE

401

show the individually tumor-specific immunogenicity of tumor-derived gp96, hsp70, hsp90, calreticulin, hsp110, and grp170 preparations has been referred to earlier (Table 1). When the hypothesis that HSPs must associate with various cellular antigens was first proposed (15, 16), it was argued broadly that HSP preparations purified from cells infected with viruses or other infectious agents must be associated with the antigenic epitopes of those agents and that such HSP-peptide complexes should be found to be immunoprotective against the cognate infectious agents (Table 1). That prediction has been amply fulfilled. Gp96 preparations isolated from influenza virus–infected cells have been shown to be protective against a challenge with the influenza virus (20, 21). Gp96 preparations isolated from mouse tissues infected with Mycobacterium tuberculosis and Listeria are protective specifically against those agents (22). A dramatic demonstration of the binding of endogenously generated antigenic peptides to gp96 came from the work of Podack and colleagues (22a), who constructed a gene encoding a gp96 molecule fused with the Fc portion of murine IgG1, generating a secretory gp96-Ig. Transfection of gp96-Ig into tumor cells decreased their tumorigenicity and increased their specific immunogenicity. The tumors were rejected after initial growth. In addition to demonstrating the binding of peptides to gp96 in vivo, these studies provide a common tool for easy generation of gp96-peptide complexes for any tumor, and possibly also for immunotherapy of human cancer. Zheng et al. (22b) have shown the broader applicability of this idea and have further developed it mechanistically, as discussed in another section. These studies do not provide definition of the antigenic epitopes associated with the HSPs, but they provide compelling circumstantial evidence by virtue of the fact that HSPs purified from antigen-negative control cells did not immunize against the particular tumor, virus, or parasite. A number of other studies provide direct evidence of association of defined antigenic epitopes with HSP molecules (Table 2). Gp96 preparations isolated from vescicular stomatitis virus (VSV)–infected or SV40-transformed cells elicit classical MHC I–restricted, antigen-specific cytotoxic T lymphocytes (CTLs) against defined antigenic epitopes of the two viruses (20, 23). In experiments with VSV, the gp96 preparations were able to cross-prime; preparations from VSV-infected cells of the b or the d haplotypes could immunize mice of the b haplotype and elicit b-specific CTLs, thus showing that gp96 was associated with peptides regardless of the MHC I haplotype of the cells from which it was purified. Arnold et al. (24) showed that immunization with gp96 preparations isolated from cells transfected with the gene encoding β-galactosidase elicited CTLs specific for an Ld-restricted epitope of β-galactosidase; similarly, immunization with gp96 preparations purified from cells expressing selected minor histocompatibility antigens was able to prime (as well as cross-prime) CTL responses against the particular minor antigens. More recently, Navaratnam et al. (25) immunized mice with gp96 isolated from cells transfected with the gD antigen of Herpes simplex virus-2. The first evidence of association of antigenic peptides with human HSPs comes from a recent study by Castelli et al. (26) who showed that human

4 Feb 2002

11:36

402

AR

AR152-14.tex

AR152-14.SGM

LaTeX2e(2001/05/10)

P1: GKD

SRIVASTAVA

melanoma-derived hsp70, but not hsp70 from other human sources, was associated with peptides corresponding to antigenic epitopes derived from gp100, Mart 1, and tyrosinase, but not Trp2. Issels and colleagues have obtained similar results with the chaperoning of tyrosinase epitopes associated with hsp70 isolated from a human melanoma (personal communication). Corresponding studies with gp96 from human melanoma cells are now in progress (C. Castelli, G. Parmiani, personal communication).

Annu. Rev. Immunol. 2002.20:395-425. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

The Structural Evidence What are the structural characteristics of the HSP-associated peptides? Immunological evidence indicates that HSPs appear to bind all peptides tested. In 13/14 instances in which the association of a given antigenic epitope with HSPs has been sought, such association has been found. (Trp2 is the sole exception thus far, and the basis of its absence in human melanoma-derived hsp70 preparations is unclear.) Based on the ability of HSP-peptide complexes to cross-prime, HSPs associate with peptides regardless of the MHC haplotype of the cells from which they are isolated. These lines of evidence point to a promiscuous ability of HSPs to bind peptides. Such promiscuity is consistent with the primordial roles of HSPs in folding and assembly of proteins, and it requires structural definition. In spite of an impressive number of studies reporting the presence of HSP-associated peptides as detected immunologically in a diverse array of systems (Table 1, 2), structural scrutiny of HSP-associated peptides has been more limited (Table 2). Four studies to date have analyzed HSP-associated peptides structurally (27–30), and each has done so with respect to a single peptide. Nieland et al. (27) first identified a known Kb-restricted viral epitope to be associated with gp96 purified from virus-infected cells; such peptides could not be detected in gp96 preparations from uninfected cells. Consistent with the cross-priming studies described earlier, the epitope was detected in VSV-infected cells of the b or the d haplotype. In a study with a mouse leukemia (28), whose Ld-restricted epitope has been defined, Ishii et al. isolated gp96, hsp90, and hsp70 from the leukemic cells. They eluted peptides from each of the three preparations and fractionated them by column chromatography. Each column fraction was tested for the presence of the antigenic epitope by pulsing Ldexpressing antigen-negative cells with it and testing the ability of specific CTLs to lyse them. Antigen-positive factions were identified among peptides eluted from each of the HSP preparations and were analyzed by mass spectroscopy. Interestingly, while each of the HSPs was associated with the precise epitope, hsp90 and gp96 were found to be associated in addition with longer precursor peptides of it. Breoler et al. (29) identified the ovalbumin-derived SIINFEKL epitope associated with gp96 and hsp70 isolated from ovalbumin-transfected cells. Meng and colleagues (30) have provided the first structural evidence for association of antigenic peptides with a human HSP. They isolated gp96 from human livers infected with hepatitis B virus and showed, by mass spectroscopy, the presence of a virus-encoded peptide with it.

4 Feb 2002

11:36

AR

AR152-14.tex

AR152-14.SGM

LaTeX2e(2001/05/10)

P1: GKD

Annu. Rev. Immunol. 2002.20:395-425. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

ROLE OF HSPs IN IMMUNE RESPONSE

403

HSP-associated peptides have not been examined thus far with the tools that have been so effective in corresponding analyses of MHC-associated peptides. Attempts in this direction have often floundered partly because the HSP molecules themselves have often degraded during elution of peptides, as a result of trace amounts of proteases present in gp96 preparations used (H-J. Schild, personal communication). However, improvements in the methods of purification used as well as the use of increasingly sophisticated mass spectroscopy tools is beginning to address this lacuna (C. Liu, P.K. Srivastava, unpublished observations). Such studies must be distinguished from others that have characterized the peptides that may be made to associate with HSPs in vitro (31, 32). Using such assays in vitro, Flynn et al. (31) suggested that “the peptide-binding site of hsp70 selects for aliphatic residues and accommodates them in an environment energetically equivalent to the interior of a folded protein.” Blond-Elguindi et al. (32) have suggested a sequence motif for peptides that may bind BiP in vitro. However, this motif is at variance with other studies that have shown hsp70-peptide binding. Blachere et al. (33) and Basu & Srivastava (34) have also analyzed a number of peptides for binding gp96, hsp70, and calreticulin in vitro and have observed considerable variation among peptides with respect to their ability to bind the HSPs. Clearly, more studies, modeled on corresponding studies with the MHC molecules, are needed to resolve the questions. The evidence for structural features of HSPs that allow them to bind peptides may be described as following. Zhu et al. (35) crystallized a ligand-binding fragment of the bacterial hsp70 known as DnaK and identified a definite peptidebinding pocket in it. “The structure consists of a beta-sandwich subdomain followed by alpha-helical segments. The peptide is bound to DnaK in an extended conformation through a channel defined by loops from the beta sandwich.” The peptide-binding activity of gp96, hsp70, and calreticulin has been demonstrated independently by Blachere et al. (33), Wearsch & Nicchitta (36), Basu & Srivastava (34) and Sastry et al. (37, 38). Wearsch et al. (36) have used fluorescent probes to identify the presence of a possible hydrophobic peptide-binding pocket in gp96. Pursuing a similar theme, Sastry et al. (37, 38) have used peptides tagged with fluorescent probes to explore the molecular environment of the peptide-binding site of gp96. Based on these studies, they have identified the amino acid position 624-630 in a highly conserved region of gp96 as the peptide-binding site (38). While these results are potentially illuminating, the not-too-well-controlled use of a large bulky probe that may alter the physicochemical properties of gp96peptide interaction places some doubt on their general validity. Interestingly, Nicchitta and colleagues as well as Sastry suggest that gp96 molecules exist as dimers and that the dimeric state is the true peptide-binding state. These observations recapitulate the original observations of Srivastava & Das (39), who demonstrated that gp96 (then called p100) molecules eluted from size exclusion columns exclusively as dimers and tetramers. The cytosolic homologue of gp96, hsp90, is also being studied structurally. Scheibel et al. (40) suggest the existence of two substrate-binding sites in hsp90, while ongoing crystal structure studies with hsp90

4 Feb 2002

11:36

Annu. Rev. Immunol. 2002.20:395-425. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

404

AR

AR152-14.tex

AR152-14.SGM

LaTeX2e(2001/05/10)

P1: GKD

SRIVASTAVA

are in the process of clarifying the identity and structure of the peptide-binding pocket of hsp90 (41–43). Buchner and colleagues (44) suggest that the hsp90 contains two and perhaps more distinct peptide-binding sites. The thesis that HSP molecules are associated with peptides including antigenic peptides was advanced solely to explain the specific immunogenicity of tumorderived homogeneous HSP preparations. Ten years later, there is overwhelming evidence for that proposition. Tumor antigens, viral antigens, antigens of intracellular parasites, mouse antigens, human antigens, cytosolic antigens, nuclear antigens, and secreted antigens have all been shown to be associated with the HSPs, and the peptide-binding pocket of at least one of the HSPs has been defined through crystallographic analysis. However, similar studies with other HSPs are yet to be carried out. The rules through which apparently any peptides are able to bind the HSPs have yet to be defined. These are decidedly rewarding avenues for future structural analyses, and the elegant and extensive work carried out with MHC-associated peptides provides a powerful precedent.

“The TSTAs of the Chemically Induced Tumors Are Still a Mystery” Let us return briefly to the initial question regarding the identity of the antigenic peptides that confer individually specific immunogenicity upon tumors. Such immunoprotective peptides have been identified in a small number of instances (see 45), and in each instance, they are mutations of normal proteins. There is no common pattern among the mutations identified, and I believe that the immunogenicity is a consequence of the random mutations that are an inevitable part of cell division (14). The individually unique antigenicity of tumors suggests a lack of relationship between the transforming and the immunogenic mutations. Unique, individually tumor-specific antigens resulting from random mutations are being increasingly identified in human cancers as well (46). With respect to the present context, the immunogenic antigens of tumors are associated with HSPs purified from the tumors in the two instances tested, i.e., a mouse leukemia (26) and a fibrosarcoma (T. Matsutake, P.K. Srivastava, submitted). Should the proposal (14) that random mutations not associated with malignant transformation are the basis of immunogenicity of mouse (and human) cancers continue to be substantiated, the mystery of TSTAs will have been resolved not in favor of an instantly gratifying molecule or family of molecules but as a myriad mutations, random but specific, in the common molecules of the cancer proteome.

HSPs Are Adjuvants: HSP-Peptide Complexes Elicit CD8+ T Cell Responses in Spite of Exogenous Administration The unequivocal demonstration that the specific immunogenicity of tumor-derived HSP preparations is elicited by HSP-associated peptides leads inevitably to the question of whether immunogenic HSP-peptide complexes could be generated in vitro. Blachere et al. (33) did just that. They reconstituted gp96-peptide and hsp70-peptide complexes in vitro using a panel of 7 peptides, and they showed

4 Feb 2002

11:36

AR

AR152-14.tex

AR152-14.SGM

LaTeX2e(2001/05/10)

P1: GKD

Annu. Rev. Immunol. 2002.20:395-425. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

ROLE OF HSPs IN IMMUNE RESPONSE

405

that while HSPs alone and peptides alone were nonimmunogenic, HSP-peptide complexes elicited MHC I–restricted antigen-specific CD8+ CTLs. These results were reproduced by Houghton (Moroi, et al. 47) who demonstrated that BiPpeptide complexes were similarly immunogenic. These authors modified the original approach of Blachere et al. (33) by using a peptide that contained the antigenic epitope and another sequence selected for a higher affinity binding to BiP. There was little evidence that the higher affinity interaction resulted in significantly higher immunogenicity over that of the unmodified peptide; however, this is an interesting question that will no doubt be resolved through future experiments. In another variation of the original approach, Suzue et al. (48) fused a mycobacterial hsp70 gene with a fragment of the ovalbumin gene and purified the fusion product. Immunization with the fusion product led to potent ovalbumin-specific CTLs and rejection of an ovalbumin-expressing tumor. A number of other genes, such as those encoding papilloma virus or malarial parasite antigens have since been fused with hsp70, and the immunogenicity of such fusion products has been demonstrated (49, 49a). In addition to settling unequivocally the questions for which they were intended, the experiments of Blachere et al. (33) led to a number of other significant findings and implications. First, they showed that HSPs could be loaded in vitro with synthetic peptides. For this, the HSPs could first be denuded of associated peptides (as by ATP treatment in case of hsp70), or they could be gently denatured in the presence of a higher temperature (50◦ C) or of guanidium hydrochloride (as in case of gp96) and then renatured in the presence of exogenous peptides. Either treatment led to association of the exogenous peptides with the HSPs. The extent of reconstitution was variable, and depending upon the HSP and the conditions used, between 1% and 10% of gp96 molecules could be loaded with peptides. Second, they showed, remarkably, that the HSP-peptide complexes were stable under conditions of denaturing polyacrylamide gel electrophoresis. Unlabeled HSPs complexed with labeled peptides migrated as radioactive bands of the size expected of the HSP-peptide complexes. This observation is consistent with our results in 1986 where a gp96 band eluted from denaturing gels was used to immunize mice successfully against the tumor from which gp96 was isolated (50). There is little precedent for this kind of noncovalent interaction, and it highlights the need for close structural examination of HSP-peptide interaction as discussed in the previous section. Third, the results of Blachere et al. show that HSP-peptide complexes elicit CD8+ T cell responses in spite of exogenous administration. Exogenous antigens are typically routed through the MHC II–presentation pathway and elicit CD4+ responses, whereas endogenously synthesized antigens are presented through MHC I molecules and stimulate CD8+ cells (51). In only a small number of instances have exogenous antigens been shown to enter the MHC I–presentation pathway (52). This demonstration makes HSPs powerful adjuvants for generation of CD8+ responses and makes them the first adjuvants of mammalian origin. (See a later section for discussion of adjuvanticity of α2 macroglobulin or α2M.) In this

4 Feb 2002

11:36

Annu. Rev. Immunol. 2002.20:395-425. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

406

AR

AR152-14.tex

AR152-14.SGM

LaTeX2e(2001/05/10)

P1: GKD

SRIVASTAVA

regard, we have recently made the not-too-surprising observation that immunization with HSP-peptide complexes also elicits antigen-specific MHC II–restricted CD4+ response (T. Matsutake and P. Srivastava, submitted for publication). Fourth, Blachere et al. (33) demonstrated that the quantity of peptide complexed to HSP molecules required for successful immunization is extremely small. As little as a few hundred picograms to a nanogram of peptide, if complexed to an HSP, was found to be sufficient to immunize. This observation, which is also germane for the adjuvanticity of the HSPs, brought into sharp focus the novelty of the mechanism of specific immunogenicity of HSP-peptide complexes (discussed in the next section). Finally, Blachere et al. (33) showed that immunogenicity did not result when peptides were complexed with mouse serum albumin that binds peptides just as effectively as the HSPs do. This suggested that the HSPs were doing something more than simply protecting the peptides from degradation or other such physical dangers. This observation too had a powerful role in our imagining of the mechanism of immunogenicity of HSP-peptide complexes.

MECHANISMS OF IMMUNOGENICITY OF HSP-PEPTIDE COMPLEXES: THE EIGHT-FOLD PATH THROUGH THE HSP RECEPTORS The Two Paths (CD8+ and CD4+) to Adaptive Immunity We were impressed with the fact that immunization with femtomole quantities of antigenic peptides chaperoned by HSPs (but not other proteins) was effective in eliciting such potent T cell responses (33). At the same time we had learned that priming of immune response by HSP-peptide complexes was exquisitely sensitive to abrogation of function of antigen presenting cells (APCs) (53). Putting these ideas together with the general biological principle that extraordinary efficiencies are often achieved through specific receptors, we proposed that HSPs interact with APCs through specific receptors and that such interaction results in endocytosis of HSP-peptide complexes followed by processing of peptides and their presentation by MHC I molecules (54). The first step in validation of this idea came from experiments that showed that macrophage, but not B cells or fibroblasts, take up gp96-peptide complexes (isolated from cells or reconstituted in vitro) and re-present the gp96-chaperoned peptides on the MHC I molecules of the macrophages; re-presentation does not occur by transfer of peptides from the gp96 molecules to MHC I on the cell surface but does require internal processing (23). Singh-Jasuja et al. (55) further demonstrated that receptor-mediated endocytosis of the gp96-chaperoned peptides is essential for re-presentation of these peptides by MHC I; nonspecific endocytosis of the gp96-peptides does not result in re-presentation. Essentially similar data for re-presentation of hsp70-chaperoned peptides were shown recently by Castellino et al. (56). The extraordinary efficiency of the process observed earlier through immunization experiments became evident again through such re-presentation

4 Feb 2002

11:36

AR

AR152-14.tex

AR152-14.SGM

LaTeX2e(2001/05/10)

P1: GKD

Annu. Rev. Immunol. 2002.20:395-425. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

ROLE OF HSPs IN IMMUNE RESPONSE

407

assays. Gp96-chaperoned peptides were re-presented by the MHC I molecules of the APC several hundred-fold more efficiently than unchaperoned peptides. Subsequently, we and others demonstrated specific, saturable, and competitive binding of various HSPs to APCs (57–59) as further evidence for the existence of an HSP receptor on APCs. In attempting to identify the long-proposed HSP receptor (54), Binder et al. (60) applied solubilized membranes of APCs on gp96 affinity columns and eluted and sequenced a gp96-binding protein. This turned out to be the previously known α2 macroglobulin (α2M) receptor CD91. α2M as well as antibodies to CD91 were shown to inhibit completely the re-presentation of gp96-chaperoned peptides by APCs. Soon thereafter, Basu et al. (61) demonstrated that CD91 acted as the receptor not only for gp96 but also for hsp90, hsp70, and calreticulin. These data were interpreted to suggest CD91 as a global HSP receptor on APCs. The wider significance of this observation is discussed later under its own heading, and we shall leave the subject of HSP receptors for now. Basu et al. (61) also shed some light on the pathway of intracellular processing of gp96-chaperoned peptides (Figure 2). It was clear from the work of Arnold et al. that the endocytosed gp96-peptide complexes enter an endosomal compartment (57) and from the work of Suto & Srivastava (23) that these compartments were not acidic. Basu et al. (61) showed that further processing of peptides required a functional proteasome and transport of the peptides through transporter associated with antigen processing (TAP), followed by the classical secretory pathway. This picture appears reasonable and straightforward. However, the mechanism of transport of peptide from the endosome to the cytosol is unclear. Further, Castellino et al. (56), who have shown an essentially similar pathway for re-presentation of hsp70-chaperoned peptides, have made the additional and remarkable observation that the structure of the peptide can dictate if the transport of the peptide into the ER is TAP-dependent or not. Thus, while the broad outlines of the internal trafficking of HSP-chaperoned peptides from the APC surface to binding to the MHC I molecules of the APCs are clear, a number of interesting and important details remain to be characterized. We have demonstrated recently that the HSP-chaperoned peptides are represented by the MHC II molecules of the APCs, in addition to re-presentation by MHC I molecules discussed thus far. Presentation of the exogenous antigens through the MHC II molecules of APCs is not surprising in and of itself. What is surprising here is the observation that the re-presentation by MHC II molecules also occurs through the CD91 receptor and that in quantitative terms, it is significantly more efficient than re-presentation through phagocytosis. This observation indicates that once an HSP-peptide complex is taken up through CD91, it may enter one or more of several trafficking and processing pathways. The factors that contribute to such molecular decisions would make for important discoveries not only for the biology of the APCs, but for cell biology in general. They will also have significant implications for strategies of vaccination. Some number crunching will illustrate how such a small quantity of antigenic peptide complexed with a small quantity of a HSP becomes powerfully immunogenic. Typically, a mouse is immunized with 1 µg of HSP-peptide

4 Feb 2002

11:36

Annu. Rev. Immunol. 2002.20:395-425. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

408

AR

AR152-14.tex

AR152-14.SGM

LaTeX2e(2001/05/10)

P1: GKD

SRIVASTAVA

Figure 2 The pathways of re-presentation of HSP-chaperoned peptides by MHC I and MHC II molecules of the APCs (macrophages and DCs) after uptake of the HSPpeptide complexes through CD91 molecules. All HSPs tested use the CD91 receptor for both pathways. The re-presentation of peptides by MHC I requires proteasomal activity and TAP (61), although TAP-independent mechanisms may also operate (56).

complexes intradermally in order to get full tumor protection. In case of gp96, this is equivalent to ∼6 × 1012 gp96 molecules chaperoning perhaps half as many peptides (assuming that the dimer is the minimum peptide-binding unit). The proportion of specific peptides among these may be estimated conservatively as 1/100,000 (in absolute numbers, 3 × 107 specific peptides in the immunizing dose). The immunization is carried out intradermally in an area of <0.1 cm2 harboring ∼104 Langerhans cells. One may imagine these 3 × 107 antigenic peptides to be delivered into the MHC I and MHC II presentation pathways of ∼104 Langerhans cells, presenting on an average of ∼3000 peptides on their MHC I and/or MHC II molecules. These cells migrate to the lymph nodes, as we have shown (62), and then stimulate the naive T lymphocytes. Ten thousand activated, mature dendritic cells (DCs) presenting 3000 MHC-peptide complexes each are a powerful

4 Feb 2002

11:36

AR

AR152-14.tex

AR152-14.SGM

LaTeX2e(2001/05/10)

P1: GKD

ROLE OF HSPs IN IMMUNE RESPONSE

409

immunological stimulus; indeed even one tenth as many DCs presenting one tenth as many antigenic peptides are powerful enough to elicit a potent T cell response.

Annu. Rev. Immunol. 2002.20:395-425. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

The Six Paths to Innate Immunity Re-presentation of HSP-chaperoned peptides by MHC molecules was first proposed and then experimentally demonstrated as a mechanism to explain the specific immunogenicity of HSP-peptide complexes. Our attention was exclusively on specificity, and we were thus surprised to find that exposure of APCs to gp96 (or other HSPs) led to secretion of low levels of TNFα by the APCs, regardless of whether or not the gp96 molecules were associated with the antigenic peptide (23). As the levels were low, and the results not consistent with our inclination, we initially attributed them to background or to contaminating lipopolysaccharides (LPS) in our HSP preparations or buffers. However, the low levels persisted reproducibly and at first look appeared unrelated to LPS contamination. This led us to a full-scale analysis of the phenomenon, the most difficult aspect of which was to convince ourselves that it was not a result of LPS contamination. With time and effort, we were able to do so and, with considerably more time and effort, were even able to publish the findings. In these reports, Basu et al. (63), Binder et al. (62), and Panjwani et al. (64) showed through in vitro and in vivo studies that the interaction of HSPs gp96, hsp90, and hsp70 with APCs led to a series of events associated with innate immunity. Singh-Jasuja et al. (65), Chen et al. (66), and Asea et al. (67) also came to partially overlapping conclusions with gp96, hsp60, and hsp70, respectively. The innate immune responses set in motion by the HSP-APC interaction may be summarized as follows: 1. Secretion of inflammatory cytokines TNFα, IL-1β, IL-12, and GM-CSF by macrophages (63). The IL-12 thus secreted acts to stimulate proliferation of NK cells (68); 2. secretion of chemokines such as MCP-1, MIP-2, and RANTES by macrophages (64, 69) and possibly by T cells (69); 3. induction of inducible nitric oxide synthase and production of nitric oxide by macrophages and DCs (64a); 4. maturation of DCs as measured by enhanced expression of MHC II, B7-2, and CD40 molecules on CD11c+ cells (63, 65); 5. migration of vast numbers of DCs (presumably Langerhans cells) from site of injection of gp96 to the draining lymph nodes (62); 6. translocation of NFκB into the nuclei of macrophages and DCs, which is perhaps the proximal event that mediates many of the other events listed here (63). Many of these phenomena have been described for murine and human macrophages and DCs. Two aspects of these studies merit specific discussion. One is

4 Feb 2002

11:36

Annu. Rev. Immunol. 2002.20:395-425. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

410

AR

AR152-14.tex

AR152-14.SGM

LaTeX2e(2001/05/10)

P1: GKD

SRIVASTAVA

the evidence that contaminating LPS is not responsible for the phenomena observed and that the activity resides in the polypeptide chain of the HSPs. This was demonstrated in several ways including an absence of measurable amounts of LPS in the HSP preparations used in the studies. Other criteria used to demonstrate the LPS-independence of the phenomena were the use of an LPS-antagonist Rslp derived from Rhodopseudomonas spheroides, the use of LPS-hypo-responsive mice, lack of requirement of LPS-binding protein (i.e., serum-independence) for the responses, and the differences in kinetics of responses from the kinetics of LPS-induced phenomena. In some studies, the heat or protease sensitivity of the phenomenon has been demonstrated as evidence of lack of involvement of LPS. Not all methods were used in all studies, but more than one method for eliminating LPS as a cause was used in most studies. The second aspect has to do with the quantity of HSPs required for elicitation of responses. Typically, HSP concentrations of 50 µg/ml to 400 µg/ml in vitro were used, although some studies were able to detect the responses with HSP concentrations as low as 1µg/ml. The higher concentrations have been criticized by some reviewers as being unphysiological. Such criticism lacks validity on two counts. First, the molar concentrations at which the HSPs are effective in eliciting the innate responses are not very different from the corresponding concentrations of LPS required to elicit the same responses. Second, the HSPs are the most abundant group of soluble proteins in cells, and if released as a result of cell lysis (see next section), they can achieve extremely high local concentrations. Thus, the APC-activating functions of HSPs can be easily imagined to occur under physiological conditions.

The Unique Position of HSPs at the Intersection of Innate and Adaptive Immune Responses The above account shows that the mechanisms through which immunization with HSP-peptide complexes elicits its potent anti-tumor, anti-viral, or anti-parasitic effect are now clear. It also illustrates how the interaction of HSPs with peptides and with APCs leads, on one hand, to stimulation of CD8+ and CD4+ T lymphocytes specific for the antigenic peptides chaperoned by the HSPs, and on the other hand, to a cascade of non-antigen-specific events (Figure 3). The latter presumably create the micro-environment necessary for effective functioning of the former. This presumption is supported by our observations (68) that while the tumorprotective activity elicited by immunization with HSP-peptide complexes could be completely abrogated by depletion of CD4+ or CD8+ T cells of the treated mice, it could be abrogated just as completely by depletion of NK cells.

CD91 AND OTHER HSP RECEPTORS ON APCS The idea of an HSP receptor is, at first sight, counterintuitive. Receptors are meant for detecting molecules that roam the body fluids and other extracellular locations. Hormones are clear examples of the kind of substances for which receptors should exist. HSPs, on the contrary, are quintessential intracellular molecules. They are

4 Feb 2002

11:36

AR

AR152-14.tex

AR152-14.SGM

LaTeX2e(2001/05/10)

P1: GKD

Annu. Rev. Immunol. 2002.20:395-425. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

ROLE OF HSPs IN IMMUNE RESPONSE

411

Figure 3 The HSP-APC interaction integrates adaptive and innate immune phenomena. It results in a series of antigen-specific and nonspecific consequences. The antigen-specific consequences (re-presentation of HSP-chaperoned peptides by the MHC I and MHC II molecules and the consequent stimulation of CD8+ and CD4+ T lymphocytes) are mediated through the CD91 HSP receptor ( filled in black). The antigen-nonspecific consequences (cytokine and chemokine release, DC maturation, etc.) are mediated by other receptors, not yet identified definitively ( filled in gray).

not found in blood, cerebrospinal fluid, synovial fluid, seminal fluid, or other secretions tested, under normal conditions (P. Srivastava, unpublished data). Why would the body need receptors for such molecules? Let us return to the reasons due to which the existence of HSP receptors was first predicted. HSP receptors were hypothesized simply to explain the mechanism for an artificial phenomenon, the extraordinary immunogenicity of HSP-peptide complexes (54). However, now that one of the HSP receptors has been identified, the physiological logic for its existence (and for existence of other HSP receptors) has come into vivid relief and sharp focus. In this section, I shall discuss the many immunological functions that have been shown to be, and others I predict will be shown to be, mediated through the HSP receptors. The evolutionary logic for the emergence of the HSP receptors also becomes strikingly clear through the studies discussed here and the thoughts that they have inspired.

α2M and the HSPs, A Certain Symmetry (or Alternatively, CD91, the Great Antigen Sampler) Identification of CD91 as an HSP receptor provides a revealing glimpse into the evolutionary origins of the HSP-APC interaction. CD91 was identified previously as a receptor for the serum protein α2M. α2M is a highly conserved molecule that

4 Feb 2002

11:36

Annu. Rev. Immunol. 2002.20:395-425. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

412

AR

AR152-14.tex

AR152-14.SGM

LaTeX2e(2001/05/10)

P1: GKD

SRIVASTAVA

can be traced as far back as Caenorhabditis elegans and perhaps earlier (70). It is also the evolutionary precursor of the C3 complement component, which in turn is the precursor for C4 and C5. I suggest in the following paragraph that in primitive organisms, α2M is involved in host defense in a manner strikingly analogous to the role of HSPs in host defense in mammals (Figure 4). Invading pathogens elaborate proteases in order to enter the host; it falls upon α2M, the protease inhibitor, to neutralize the proteases and foil the pathogen’s designs. Further, the α2M is a protease inhibitor of a most unusual sort; the molecules contain a “bait” region that harbors sequences that are substrates for selected proteases. Once the proteases take the bait, i.e., proteolyse it, the α2M molecule acquires an altered conformation that physically traps the proteases into a molecular basket. The protease, now trapped

Figure 4 The author’s view of how a mechanism of nonspecific surveillance against external pathogens in primitive organisms evolved into an integral part of a complex adaptive immune response to internal pathogens and cancers. Evolution of a mechanism of indirect presentation is inherent in the scheme. (A) Inhibition, uptake, and degradation of protease secreted by an external pathogen by the host α2M/receptor on phagocytes of primitive organisms. (B) Degradation of internal pathogen, binding of the peptides by HSPs, lysis of host cell, release of HSP-peptides, their uptake by host α2M/receptor CD91, and re-presentation by MHC I molecules of the vertebrate APC. Those symbols not identified here are the same as in Figures 2 and 3.

4 Feb 2002

11:36

AR

AR152-14.tex

AR152-14.SGM

LaTeX2e(2001/05/10)

P1: GKD

Annu. Rev. Immunol. 2002.20:395-425. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

ROLE OF HSPs IN IMMUNE RESPONSE

413

by α2M, is endocytosed through the α2M receptor CD91 and internalized into the phagocytes, wherein the foreign protease is digested by the host proteases. The digested products, the amino acids, act as a source of nutrition for the host. Thus, α2M combines the functions of host defense and nutrition. Now, let us fast forward the evolutionary videotape by several hundred million years. The landscape has changed substantially. The pathogens have now acquired sophisticated means of entry into the host other than secreting proteases. Viruses and intracellular parasites have developed ligands for host receptors such that they can enter the host formally through proper channels rather than by breaking down the walls through proteases. Having entered the host cells, they can now establish residence there and replicate. What is the α2M to do? It is a secreted protein selected to patrol the periphery, capture the enemy, and bring it in through the CD91 portal to be digested. It is not trained to scan the intracellular environment. Lo and behold! There has always existed a potent and abundant intracellular agent that happens to come into contact with the entire intracellular contents (including but not limited to invading pathogens) as it helps fold the proteins, chaperones them to interact with other proteins, and helps degrade them when necessary. It is the HSPs. They carry bits and pieces of everything they encounter. They are now co-opted to play their role in defense. The next steps are simple. The pathogens infect the cells, replicate there, and at some point destroy them. The abundant HSPs carrying the information about the internal environment are released and now dock to the CD91 that has by now developed a far more sophisticated downstream mechanism than simply digesting the enemy (the nutritional needs of the host now being met through an independent agency). The new mechanisms have been selected because the enemy may now look very much like the host; to tell one from the other, a complex and independent adaptive immune system has developed. The CD91 along with the HSPs carrying their information (peptides) is internalized; while some of it still goes to an acidic compartment for digestion of the contents as in the days of yore, the rest of it now goes to a separate nonacidic compartment that channels the peptides through a complex processing pathway that leads to their presentation by the MHC I molecules of the phagocyte (now called the APC). The MHC I molecules simply bind what they can and present them to the T cells, which alone can tell if a given peptide hints at the presence of an enemy. If it does, the T cells expand and go on a search and destroy mission. This is my personal view of how the surveillance of the external milieu by the α2M/CD91 system came to be transformed into an efficient mechanism of internal surveillance by the HSP/CD91 system (Figure 4). A view of the evolution of the mechanism of indirect presentation is inherent in this opinion. The primordial functions of HSPs as chaperones came to be recruited into the needs of an increasingly complex security apparatus whose challenges had transformed from protecting the host from pathogens that were bluntly violating the host’s perimeters to those, i.e., viruses and intracellular parasites, that had learned the ways of the hosts and learned to live within them. The symmetry between α2M and HSPs

4 Feb 2002

11:36

Annu. Rev. Immunol. 2002.20:395-425. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

414

AR

AR152-14.tex

AR152-14.SGM

LaTeX2e(2001/05/10)

P1: GKD

SRIVASTAVA

becomes evident in this view. This view also makes the CD91 the central antigenic sampling device for both the extracellular and the intracellular antigenic worlds, and this view is further elaborated in a subsequent section where I argue that the HSP receptors, including but not limited to CD91, act as sensors of necrotic cell death and more broadly as sensors of cellular stress. The symmetry between α2M and HSPs has an interesting practical application. Binder et al. (71) argued that as α2M and HSPs use the same CD91 portal into the APCs and as HSP-peptide complexes entering through this portal are presented through the MHC I molecules, peptides chaperoned by α2M should be similarly processed and presented by MHC I molecules. The α2M-peptide complexes should therefore be effective immunogens as well. This is indeed the case, further demonstrating an additional symmetry between HSPs and α2M.

CD91: Key Portal for Indirect Presentation or Cross-Priming? CD91 has been shown by Basu et al. to be a common receptor for gp96, hsp90, hsp70, and calreticulin (61). Our ongoing studies indicate that it may indeed be a common receptor for other HSPs as well (R. Binder, P.K. Srivastava, unpublished). Basu et al. make the interesting observation that blocking of CD91 through an antibody to it, or through its ligand α2M, can inhibit completely the phenomenon of re-presentation of peptides chaperoned by any of the HSPs tested. This result indicates not only that CD91 is a receptor for the HSPs tested, but also that it is the sole receptor for them with respect to their peptide re-presenting function. This observation has an important implication for one of the most significant immunological phenomena, indirect presentation or cross-priming. Suto & Srivastava (23) and Arnold et al. (24) have shown previously that HSP preparations can cross-prime, i.e., gp96 preparations from an antigen-positive cell of a given MHC haplotype can immunize mice of a different haplotype and elicit CTL responses restricted by the MHC I of the other haplotype. Our ongoing studies (R. Binder, P.K. Srivastava, unpublished) now indicate that not only gp96, but other HSPs, i.e., hsp70, hsp90, and calreticulin as well, can cross-prime. They further indicate that if HSPs are rendered unavailable, indirect presentation of an antigen cannot occur and that HSPs are necessary for cross-priming or indirect presentation. In other words, in order for an antigen to be re-presented, it must exist in a form that is complexed with one or more HSPs. To the extent that the results of this study are valid and generalizable, it follows that CD91 is a key portal for indirect presentation or cross-priming. In addition to the experiments discussed above, quantitative considerations propel us into pursuing the idea that HSP-antigen complexes are the preferred form for channeling antigens into the pathway for indirect presentation in vivo. Under physiological conditions, indirect presentation requires tremendous quantitative economy with respect to the antigen. Typically, a small number of cells may be infected by a virus or parasite, and the antigens involved are not necessarily expressed in abundant quantities. Indeed, there is little correlation between the abundance of an antigen and its re-presentability, suggesting that the mechanism that mediates

4 Feb 2002

11:36

AR

AR152-14.tex

AR152-14.SGM

LaTeX2e(2001/05/10)

P1: GKD

Annu. Rev. Immunol. 2002.20:395-425. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

ROLE OF HSPs IN IMMUNE RESPONSE

415

indirect presentation in vivo is able to do so with very little antigen. This in turn suggests the involvement of receptor-mediated mechanisms. Apart from its biological common sense, this line of reasoning is supported by Lanzaveccia’s demonstration (72) that the uptake of antigen for re-presentation by MHC II molecules is more efficient by several orders of magnitude if mediated by a receptor-mediated mechanism (antigen-specific B lymphocytes) than by a receptor-independent mechanism. Of the few mechanisms that have been claimed to be candidates for indirect presentation (12), only two are receptor-mediated, the CD91 HSP receptor-mediated antigen uptake and the uptake of antibody-bound antigens through the Fc receptor (73). Without negating the important role that the latter mechanism must play under certain conditions, HSP-mediated antigen uptake is particularly attractive, as HSPs have been shown to chaperone antigenic peptides from all cellular compartments and of a wide variety, including tumor antigens, viral antigens, minor histocompatibility antigens, and model antigens as discussed earlier. The HSPs thus appear to be a universal mechanism for antigen-capture, and they permit a high-efficiency antigen uptake through a receptor-mediated mechanism. While much of the excitement about cross-priming has centered around stimulation of CD8+ T cells by MHC I–peptide complexes for obvious reasons, crosspriming also occurs just as readily for MHC II–presented antigens. On the basis of quantitative considerations discussed above, it is my belief that cross-priming for this pathway also occurs through HSP-peptide complexes, although there is no mechanistic necessity for it.

Other HSP Receptors as Key Players in Innate Immunity The discussion of HSP receptors has focused thus far on the receptor involved in representation of HSP-chaperoned peptides, CD91. This is obviously asymmetrical because the HSP-APC interaction has many significant consequences other than representation of HSP-chaperoned peptides. These include the antigen-independent elaboration of cytokines and chemokines and translocation of NFκB into the nucleus and other effects previously discussed. There is little evidence that CD91 is the receptor involved in these phenomena. Considering that the other phenomena must involve signal transduction, and that it is not clear if CD91 is a signaling receptor, there is a strong likelihood that other receptors are involved. Indeed, there is some evidence for a role for other receptors. Ohashi et al. have implicated the LPS receptor tlr4 in signaling by hsp60 (74), and Panjwani et al. believe CD36 to be a signaling receptor for gp96 (75). These studies are still quite preliminary and have not established that the HSPs actually interact physically with the tlr4 or CD36, nor have these studies established the downstream signaling pathways involved. The lack of maturity of these results notwithstanding, it is reasonable that the identity of these or other receptors, the molecular details of their interactions with HSPs, and the signaling cascades initiated by such interactions will be revealed in the near future. These other putative receptors are most likely to be involved in antigen-independent mechanisms and thus are expected to be the key

4 Feb 2002

11:36

Annu. Rev. Immunol. 2002.20:395-425. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

416

AR

AR152-14.tex

AR152-14.SGM

LaTeX2e(2001/05/10)

P1: GKD

SRIVASTAVA

players in the innate immune responses initiated by HSP-APC interaction. Since IL-1-like activities are secreted by phagocytes of the most primitive multicellular organisms, the HSP receptor(s) on APCs that mediate secretion of IL-1 are likely to provide the primordial template of a signaling circuit involved in innate immunity. The discussion in this section has thus far centered on the HSP-APC interaction. Newly emerging evidence indicates that HSPs may interact with cells other than APCs, such as platelets (76), NK cells (77), and T cells (69, 78). These observations are still relatively preliminary and will no doubt reveal novel aspects of the role of HSPs in innate immunity.

HSP Receptors as Sensors of Necrotic Cell Death The paradox that HSPs are intracellular molecules and yet CD91 and other receptors for them exist has been discussed at the very beginning of this section. The resolution to the paradox lies within it: HSP receptors make perfect sense because HSPs are quintessentially intracellular and abundant soluble molecules under normal conditions. If extracellular HSPs are detected by a receptor through release of cytokines or chemokines, or maturation and migration in the case of DCs, the conditions must be abnormal. The recent demonstration by Basu et al. (63) that hsp70, hsp90, gp96, and calreticulin are released from cells as a result of necrotic but not apoptotic death is germane in this regard. Nicchitta and colleagues have similarly observed that gp96 is released from cells undergoing virus-induced lytic death but not from cells dying apoptotically (79). Melcher et al. (80) had reported earlier in a pioneering study that tumor cells undergoing necrotic death are highly immunogenic as compared with those undergoing apoptotic death. Necrotic death may not be a physiological event, whereas vast numbers of cells die apoptotically during embryonic develoment, thymic selection, and other processes. Thus, the detection of HSPs by a receptor is an excellent mechanism to signal an abnormal loss of physical integrity. Considering the phylogenetic antiquity of phagocytes and of HSPs, it is safe to suggest that the release of HSPs due to inappropriate necrotic death may be an ancient mechanism for making a host aware that bad things were happening to it. The recent demonstration by Galucci et al. (81) and Sauter et al. (82) that necrotic cell lysates but not apoptotic cells cause maturation of DCs is consistent with these ideas. I believe that HSPs are the major components of the DC-maturing activity of the necrotic lysates, although other components such as DNA also cause activation of APCs (83). These ideas may be generalized to say that necrotic cell death may not be the only event that exposes APCs to HSPs. Stressed cells and cancer cells have been reported to express cell surface HSP molecules (50, 77, 84, 85), and it is tempting to imagine that such cells activate APCs directly. Interestingly, Li and colleagues have observed recently that physical contact of tumor cells artificially engineered to express cell surface HSPs with immature DCs elicits a powerful maturation of DCs (22b). Bhardwaj and colleagues have recently observed that murine and human cancer cells show elevated levels of hsp70 and gp96, and their necrotic lysates have elevated DC-maturing activity (85a). These observations blend seamlessly with the early observation of Menoret et al. (86) and

4 Feb 2002

11:36

AR

AR152-14.tex

AR152-14.SGM

LaTeX2e(2001/05/10)

P1: GKD

Annu. Rev. Immunol. 2002.20:395-425. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

ROLE OF HSPs IN IMMUNE RESPONSE

417

Melcher et al. (80) who demonstrated that the immunogenicity of tumors cosegregated with expression of inducible but not constitutive hsp70. Vanaja et al. (86a) and Clark & Menoret (87) have shown recently that heat shocked tumor cells that express high levels of inducible HSPs are more immunogenic than their constitutive counterparts. Further exploration of these ideas has now lead Callahan et al. (88) to explore the differences in peptide-binding abilities of inducible versus constitutive hsp70. Whether the inducible HSPs have unique effects on APCs, distinct from the effects of constitutive HSPs, is a fascinating and biologically attractive idea awaiting further experimental attention. We integrated these ideas and suggested recently that the HSP receptor CD91 is a sensor of necrotic cell death. In view of the recognition that there might be two classes of HSP receptors—CD91 involved in re-presentation and others (possibly tlr4, CD36) involved in the innate components of APC activation—we extend our previous suggestion to say that HSP receptors as a class are sensors of necrotic cell death. This discussion will not be complete without a mention of the elegant work of Fadok et al. (89) who have identified, through an experimental tour de force, a receptor for phosphatidyl serine, a marker for apoptotic cells, on APCs. They further showed that engagement of this receptor activates the anti-inflammatory program in APCs. This phenomenon is the mirror image of our observations that activation of HSP receptors (read, necrosis receptors) activates the pro-inflammatory program in APCs. The interesting possibility must be considered that the HSPs and phosphatidyl serine inversely modulate the APC receptors for the other as a mechanism to maximize their effect on the APCs.

Evolutionary Conservation of HSP-APC Interaction Following Dobzhansky’s dictum that biological phenomena could not be understood unless viewed through the prism of evolution, we sought evidence for immunogenicity of HSP-peptide complexes in the earliest vertebrate that has transplantable tumors and has an adaptive immune system. Inevitably, we arrived at the Xenopus. Using a syngeneic transplantable tumor of the Xenopus, Robert et al. (90) showed the specific immunoprotective activity of the tumor-derived hsp70 and gp96 against the tumor from which the HSPs were purified. They also showed the dependence of this phenomenon on the presence of HSP-associated peptides. In a cross-cultural experiment, the APCs of mice pulsed with Xenopus HSP complexed in vitro with a mouse CTL epitope were able to re-present the Xenopus HSP-chaperoned peptide to mouse CTLs. In so doing, we were able to recapitulate in the phylogenetically distant Xenopus, the essential elements of HSP-peptide and HSP-APC interaction described thus far in mice, rats, and humans.

HSPS AND MHC: A CIRCLE CLOSED The common experimental pedigree of the discovery of MHC molecules and of the immunological functions of HSPs was alluded to in the beginning of this chapter. In closing, I would like to comment on the continuing intertwined roles of these

4 Feb 2002

11:36

Annu. Rev. Immunol. 2002.20:395-425. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

418

AR

AR152-14.tex

AR152-14.SGM

LaTeX2e(2001/05/10)

P1: GKD

SRIVASTAVA

two groups of molecules. The MHC and the HSPs are peptide-binding proteins, although the association is far more promiscuous in the case of one than the other. The differences in the levels of promiscuity have an evolutionary origin and a physiological consequence. The HSPs began to bind peptides at the very beginning of life, as a collateral consequence of their general chaperoning functions. The MHC proteins, on the other hand, were perhaps selected for peptide-binding in an evolutionary era of increasing organismal complexity. The differences in the stringency of requirements for peptide-binding by the two molecules make it possible for one (the HSPs) to take all comers and to render them presentable by the other (the MHC), depending on the latter’s more discerning demands. From these differences arises the ability of HSPs (and I believe, an essential function of HSPs) to cross-prime. The innate ability of HSPs to scan the entire protein repertoire of proteins inside the cells, also as a collateral to their normal chaperoning function, allows them to play a key role as informers of the MHC molecules (and through them of the T cells) with respect to any pathogens that might lurk inside the host. The ability of HSPs to interact specifically with the APCs through the HSP receptors, selected I believe as early as the appearance of the phagocyte, imparted on them the ability to be economical with the precious quantities of antigens to be presented to the MHC molecules. Finally, we have long suggested (54) that the HSPs of the cytosol and the ER play a role in chaperoning the peptides from the point of their generation to the point of their being loaded onto the MHC I molecules, and indeed HSPs help process them through their inherent aminopeptidase activity (91). These are some of the many ways in which I believe that the roles of these two molecules are essentially interwoven to form the majestic tapestry of immune response. We took these ideas far, perhaps too far, and suggested in 1991 that the lack of any obvious sequence homologies between the MHC and the HSPs notwithstanding, the two molecules may have common evolutionary precursors. The emerging structural evidence with respect to hsp70 does not support that notion (35), although the gp96 molecule has recently been modeled on the basis of the MHC I molecule (38). Regardless of what the structures tell us in the future, the functional convergence of the MHC and the HSPs is increasingly apparent and esthetically appealing; esthetic appeal remains an undeniably excellent basis on which to form experimentally testable hypotheses. This is an appropriate place to comment on some of the experimental avenues that have been so productive in exploration of structure and function of the MHC and that are not available for the study of the HSPs. The lack of polymorphisms and of naturally arising HSP mutants, the consequent inability to do genetics, and to extinguish expression of HSPs in cells in culture or in vivo due to cellular or embryonic lethality are some of the obvious experimental limitations that result from the vital roles of the HSPs in nonimmunological processes. The lack of these tools has made it difficult to pin down many of the roles of HSPs in immune response with the finality that one would like. Evidence for many of the phenomena described here therefore was circumstantial rather than direct, the more so when the observations or the interpretations were first made. Nonetheless, the observations

4 Feb 2002

11:36

AR

AR152-14.tex

AR152-14.SGM

LaTeX2e(2001/05/10)

P1: GKD

ROLE OF HSPs IN IMMUNE RESPONSE

419

have firmly stood the test of independent reproducibility, and the interpretations and predictions have, without exception, proven to be correct. Altogether, the evidence for a fundamental physiological role of HSPs in an array of immunological phenomena continues to widen.

Annu. Rev. Immunol. 2002.20:395-425. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

THE ROADS NOT TAKEN This area of work began with exploration of the biochemical basis of immunogenicity of tumors (39). It would have been appropriate to end it with a discussion of the clinical applications of the ideas that have come out of this exploration. However, space does not permit that description, nor does it permit a discussion of the applications of the HSP approach for immunotherapy of infectious diseases. The reader is referred to other reviews for these purposes (45, 92). Space also does not permit discussion of the dose-restriction of immunogenicity of HSP-peptide complexes and its implications for generating downregulatory antigen-specific T cells (93, 94). Finally, the idea that HSPs chaperone antigenic peptides from the point of their generation in the cytosol to their being loaded onto MHC I molecules in the ER, proposed some time ago (54), and an idea that has been as hard to prove as it has been to disprove, but which continues to gather support (91, 95–97a), is not discussed here. These omissions are solely for reasons of limitations of space. Visit the Annual Reviews home page at www.annualreviews.org

LITERATURE CITED 1. Lindquist S, Craig EA. 1988. The heatshock proteins. Annu. Rev. Genet. 22:631– 77 2. Ritossa F. 1962. A new puffing pattern induced by temperature shock and DNP in Drosophila. Experientia 18:571–73 3. Gething MJ, Sambrook J. 1992. Protein folding in the cell. Nature 355:33–45 4. Parsell DA, Lindquist S. 1993. The function of heat-shock proteins in stress tolerance: degradation and reactivation of damaged proteins. Annu. Rev. Genet. 27: 437–96 5. Haas IG. 1991. BiP—a heat shock protein involved in immunoglobulin chain assembly. Curr. Top. Microbiol. Immunol. 167:71–82 6. Lindquist S. 1986. The heat-shock response. Annu. Rev. Biochem. 55:1151–91 7. Rutherford SL, Lindquist S. 1998. Hsp90

8.

9.

10.

11.

as a capacitor for morphological evolution. Nature 396:336–42 Morimoto RI. 1998. Regulation of the heat shock transcriptional response: cross talk between a family of heat shock factors, molecular chaperones, and negative regulators. Genes Dev. 12(24):3788–96 Feder ME, Hofmann GE. 1999. Heatshock proteins, molecular chaperones, and the stress response: evolutionary and ecological physiology. Annu. Rev. Physiol. 61:243–82 Klein G. The strange road to the tumorspecific transplantation antigens (TSTAs). Cancer Immunity 1:6 Klein G, Sjorgen HO, Klein E, Hellstrom KE. 1960. Demonstration of resistance against methylcholanthrene-induced sarcomas in the primary autochthonous host. Cancer Res. 20:1561–72

4 Feb 2002

11:36

Annu. Rev. Immunol. 2002.20:395-425. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

420

AR

AR152-14.tex

AR152-14.SGM

LaTeX2e(2001/05/10)

P1: GKD

SRIVASTAVA

12. Srivastava PK, Menoret A, Basu S, Binder RJ, McQuade KL. 1998. Heat shock proteins come of age: primitive functions acquire new roles in an adaptive world. Immunity 8(6):657–65 13. Udono H, Srivastava PK. 1994. Comparison of tumor-specific immunogenicities of stress-induced proteins gp96, hsp90, and hsp70. J. Immunol. 152:5398–403 14. Srivastava PK. 1993. Peptide-binding heat shock proteins in the endoplasmic reticulum: role in immune response to cancer and in antigen presentation. Adv. Cancer Res. 62:153–77 15. Srivastava PK, Heike M. 1991. Tumorspecific immunogenicity of stress-induced proteins: convergence of two evolutionary pathways of antigen presentation? Semin. Immunol. 3:57–64 16. Srivastava PK, Maki RG. 1991. Stressinduced proteins in immune response to cancer. Curr. Top. Microbiol. Immunol. 167:109–23 17. Li Z, Srivastava PK. 1993. Tumor rejection antigen gp96/grp94 is an ATPase: implications for protein folding and antigen presentation. EMBO J. 12:3143–51 18. Udono H, Srivastava PK. 1993. Heat shock protein 70-associated peptides elicit specific cancer immunity. J. Exp. Med. 178:1391–96 19. Menoret A, Peng P, Srivastava PK. 1999. Association of peptides with heat shock protein gp96 occurs in vivo and not after cell lysis. Biochem. Biophys. Res. Commun. 262(3):813–18 20. Blachere NE. 1998. The role of heat shock proteins in antigen presentation by MHC I molecules. Ph.D thesis, Fordham Univ., New York 21. Heikema A, Agsteribbe E, Wiscjut J, Huckriede A. 1997. Generation of heat shock protein-based vaccines by intracellular loading of gp96 with antigenic peptides. Immunol. Lett. 57:69–74 22. Zugel U, Sponaas AM, Neckermann J, Schoel B, Kaufmann SH. 2001. gp96peptide vaccination of mice against

22a.

22b.

23.

24.

25.

26.

27.

28.

29.

intracellular bacteria. Infect. Immunol. 69(6):4164–67 Yamazaki K, Nguyen T, Podack ER. 1999. Cutting edge: Tumor secreted heat shock-fusion protein elicits CD8 cells for rejection. J. Immunol. 163(10):5178–82 Zheng H, Dai J, Stoilova D, Li Z. 2002. Surface targeting of an intracellular heat shock protein gp96 on tumor cells induces dendritic cell maturation and anti-tumor immunity. J. Immunol. In press Suto R, Srivastava PK. 1995. A mechanism for the specific immunogenicity of heat shock protein-chaperoned peptides. Science 269:1585–88 Arnold D, Faath S, Rammensee H, Schild H. 1995. Cross-priming of minor histocompatibility antigen-specific cytotoxic T cells upon immunization with the heat shock protein gp96. J. Exp. Med. 182: 885–88 Navaratnam M, Deshpande MS, Hariharan MJ, Zatechka DS, Srikumaran S. 2001. Heat shock protein-peptide complexes elicit cytotoxic T-lymphocyte and antibody responses specific for bovine herpesvirus 1. Vaccine 19(11-12):1425– 34 Castelli C, Ciupitu AM, Rini F, Rivoltini L, Mazzocchi A, Kiessling R, Parmiani G. 2001. Human heat shock protein 70 peptide complexes specifically activate antimelanoma T cells. Cancer Res. 61(1):222–27 Nieland TJ, Tan MC, Monne-van Muijen M, Koning F, Kruisbeek AM, van Bleek GM. 1996. Isolation of an immunodominant viral peptide that is endogenously bound to the stress protein GP96/GRP94. Proc. Natl. Acad. Sci. USA 93:6135–39 Ishii T, Udono H, Yamano T, Ohta H, Uenaka A, Ono T, Hizuta A, Tanaka N, Srivastava PK, Nakayama E. 1999. Isolation of MHC class I-restricted tumor antigen peptide and its precursors associated with heat shock proteins hsp70, hsp90, and gp96. J. Immunol. 162(3):1303–9 Breloer M, Marti T, Fleischer B, von

4 Feb 2002

11:36

AR

AR152-14.tex

AR152-14.SGM

LaTeX2e(2001/05/10)

P1: GKD

ROLE OF HSPs IN IMMUNE RESPONSE

30.

Annu. Rev. Immunol. 2002.20:395-425. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

31.

32.

33.

34.

35.

36.

37.

Bonin A. 1998. Isolation of processed, H-2Kb-binding ovalbumin-derived peptides associated with the stress proteins HSP70 and gp96. Eur. J. Immunol. Mar 28(3):1016–21 Meng SD, Gao T, Gao GF, Tien P. 2001. HBV-specific peptide associated with heat-shock protein gp96. Lancet 357(9255):528–29 Flynn GC, Pohl J, Flocco MT, Rothman JE. 1991. Peptide-binding specificity of the molecular chaperone BiP. Nature 353(6346):726–30 Blond-Elguindi S, Cwirla SE, Dower WJ, Lipshutz RJ, Sprang SR, Sambrook JF, Gething MJ. 1993. Affinity panning of a library of peptides displayed on bacteriophages reveals the binding specificity of BiP. Cell 75(4):717–28 Blachere NE, Li Z, Chandawarkar RY, Suto R, Jaikaria NS, Basu S, Udono H, Srivastava PK. 1997. Heat shock proteinpeptide complexes, reconstituted in vitro, elicit peptide-specific cytotoxic T lymphocyte response and tumor immunity. J. Exp. Med. 186:1315–22 Basu S, Srivastava PK. 1999. Calreticulin, a peptide-binding chaperone of the endoplasmic reticulum, elicits tumor- and peptide-specific immunity. J. Exp. Med. 189(5):797–802 Zhu X, Zhao X, Burkholder WF, Gragerov A, Ogata CM, Gottesman ME, Hendrickson WA. 1996. Structural analysis of substrate binding by the molecular chaperone DnaK. Science 272:1606–14 Wearsch PA, Nicchitta CV. 1997. Interaction of endoplasmic reticulum chaperone GRP94 with peptide substrates is adenine nucleotide-independent. J. Biol. Chem. 272:5152–56 Linderoth NA, Simon MN, Rodionova NA, Cadene M, Laws WR, Chait BT, Sastry S. 2001. Biophysical analysis of the endoplasmic reticulum-resident chaperone/heat shock protein gp96/GRP94 and its complex with peptide antigen. Biochem. 40(5):1483–95

421

38. Linderoth NA, Popowicz A, Sastry S. 2000. Identification of the peptide-binding site in the heat shock chaperone/tumor rejection antigen gp96 (Grp94). J. Biol. Chem. 275(8):5472–77 39. Srivastava PK, Das MR. 1984. Serologically unique surface antigen of a rat hepatoma is also its tumor-associated transplantation antigen. Int. J. Cancer 33: 417–22 40. Scheibel T, Weikl T, Buchner J. 1998. Two chaperone sites in Hsp90 differing in substrate specificity and ATP dependence. Proc. Natl. Acad. Sci. USA 95:1495–99 41. Stebbins CE, Russo AA, Rosen N, Hartl FU, Pavletich NP. 1997. Crystal structure of an Hsp90-geldanamycin complex: targeting of a protein chaperone by an antitumor agent. Cell 89:239–50 42. Prodromou C, Roe SM, Piper PW, Pearl LH. 1997. A molecular clamp in the crystal structure of the N-terminal domain of the yeast hsp90-chaperone. Nat. Struct. Biol. 4:477–82 43. Prodromou C, Roe SM, Obrien R, Ladbury JE, Piper PW, Pearl LH. 1997. Identification and structural characterization of the AYP/ADP-binding site in the hsp90 molecular chaperone. Cell 90:65–70 44. Scheibel T, Weikl T, Buchner J. 1998. Two chaperone sites in Hsp90 differing in substrate specificity and ATP dependence. Proc. Natl. Acad. Sci. USA 95(4):1495– 99 45. Srivastava PK, Kumar S, Mendonca C. 2001. Principles and practice of the use of heat shock protein-peptide complexes for immunotherapy of human cancer. Prin. Practice Biol. Ther. Cancer Updates 2(3):1–11 46. Van den Eynde B, van der Bruggen P. 2001. T-cell defined tumor antigens. Cancer Immunity (www.cancerimmunity. org/peptidedatabase/Tcellepitopes.htm) 47. Moroi Y, Mayhew M, Trcka J, Hoe MH, Takechi Y, Hartl FU, Rothman JE, Houghton AN. 2000. Induction of cellular immunity by immunization with novel

4 Feb 2002

11:36

422

48.

Annu. Rev. Immunol. 2002.20:395-425. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

49.

49a.

50.

51.

52.

53.

54.

55.

AR

AR152-14.tex

AR152-14.SGM

LaTeX2e(2001/05/10)

P1: GKD

SRIVASTAVA hybrid peptides complexed to heat shock protein 70. Proc. Natl. Acad. Sci. USA 97(7):3485–90 Suzue K, Zhou X, Eisen HN, Young RA. 1997. Heat shock fusion proteins as vehicles for antigen delivery into the major histocompatibility complex class I presentation pathway. Proc. Natl. Acad. Sci. USA 94(24):13146–51 Cheng WF, Hung CF, Chai CY, Hsu KF, He L, Rice CM, Ling M, Wu TC. 2001. Enhancement of Sindbis virus self-replicating RNA vaccine potency by linkage of Mycobacterium tuberculosis heat shock protein 70 gene to an antigen gene. J. Immunol. 166(10):6218–26 Udono H, Yamano T, Kawabata Y, Ueda M, Yui K. 2001. Generation of cytotoxic T lymphocytes by MHC class I ligands fused to heat shock cognate protein 70. Int. Immunol. 13(10):1233–42 Srivastava PK, DeLeo AB, Old LJ. 1986. Tumor rejection antigens of chemically induced sarcomas of inbred mice. Proc. Natl. Acad. Sci. USA 83:3407–11 Townsend A, Bodmer H. 1989. Antigen recognition by class I-restricted T lymphocytes. Annu. Rev. Immunol. 7:601–24 Yewdell JW, Norbury CC, Bennink JR. 1999. Mechanisms of exogenous antigen presentation by MHC class I molecules in vitro and in vivo: implications for generating CD8+ T cell responses to infectious agents, tumors, transplants, and vaccines. Adv. Immunol. 73:1–77 Udono H, Levey DL, Srivastava PK. 1994. Cellular requirements for tumorspecific immunity elicited by heat shock proteins: tumor rejection antigen gp96 primes CD8+ T cells in vivo. Proc. Natl. Acad. Sci. USA 91:3077–81 Srivastava PK, Udono H, Blachere NE, Li Z. 1994. Heat shock proteins transfer peptides during antigen processing and CTL priming. Immunogenetics 39:93–98 Singh-Jasuja H, Toes RE, Spee P, Munz C, Hilf N, Schoenberger SP, RicciardiCastagnoli P, Neefjes J, Rammensee HG,

56.

57.

58.

59.

60.

61.

62.

63.

Arnold-Schild D, Schild H. 2000. Crosspresentation of glycoprotein 96-associated antigens on major histocompatibility complex class I molecules requires receptor-mediated endocytosis. J. Exp. Med. 191(11):1965–74 Castellino F, Boucher PE, Eichelberg K, Mayhew M, Rothman JE, Houghton AN, Germain RN. 2000. Receptor-mediated uptake of antigen/heat shock protein complexes results in major histocompatibility complex class I antigen presentation via two distinct processing pathways. J. Exp. Med. 191(11):1957–64 Arnold-Schild D, Hanau D, Spehner D, Schmid C, Rammensee HG, de la Salle H, Schild H. 1999. Cutting edge: receptormediated endocytosis of heat shock proteins by professional antigen-presenting cells. J. Immunol. 162(7):3757–60 Wassenberg JJ, Dezfulian C, Nicchitta CV. 1999. Receptor mediated and fluid phase pathways for internalization of the ER Hsp90 chaperone GRP94 in murine macrophages. J. Cell Sci. 112(Pt 13):2167–75 Binder RJ, Harris ML, Menoret A, Srivastava PK. 2000. Saturation, competition, and specificity in interaction of heat shock proteins (hsp) gp96, hsp90, and hsp70 with CD11b+ cells. J. Immunol. 165(5):2582–87 Binder RJ, Han DK, Srivastava PK. 2000. CD91: a receptor for heat shock protein gp96. Nat. Immunol. 1(2):151–55 Basu S, Binder RJ, Ramalingam T, Srivastava PK. 2001. CD91 is a common receptor for heat shock proteins gp96, hsp90, hsp70, and calreticulin. Immunity 14(3):303–13 Binder RJ, Anderson KM, Basu S, Srivastava PK. 2000. Cutting edge: Heat shock protein gp96 induces maturation and migration of CD11c+ cells in vivo. J. Immunol. 165(11):6029–35 Basu S, Binder RJ, Suto R, Anderson KM, Srivastava PK. 2000. Necrotic but not apoptotic cell death releases heat

4 Feb 2002

11:36

AR

AR152-14.tex

AR152-14.SGM

LaTeX2e(2001/05/10)

P1: GKD

ROLE OF HSPs IN IMMUNE RESPONSE

64.

Annu. Rev. Immunol. 2002.20:395-425. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

64a.

65.

66.

67.

68.

69.

70.

shock proteins, which deliver a partial maturation signal to dendritic cells and activate the NF-kappa B pathway. Int. Immunol. 12(11):1539–46 Panjwani NN, Popova L, Srivastava PK. 2000. Production of chemokines by hsp70 and gp96 stimulated antigen presenting cells. Cell Stress Chaperones 5:392 Panjwani NN, Popova L, Srivastava PK. 2002. Heat shock proteins gp96 and hsp70 activate release of nitric oxide by antigen presenting cells. J. Immunol. In press Singh-Jasuja H, Scherer HU, Hilf N, Arnold-Schild D, Rammensee HG, Toes RE, Schild H. 2000. The heat shock protein gp96 induces maturation of dendritic cells and down-regulation of its receptor. Eur. J. Immunol. 30(18):2211–15 Chen W, Syldath U, Bellmann K, Burkart V, Kolb H. 1999. Human 60-kDa heat-shock protein: a danger signal to the innate immune system. J. Immunol. 162(6):3212–19 Asea A, Kraeft SK, Kurt-Jones EA, Stevenson MA, Chen LB, Finberg RW, Koo GC, Calderwood SK. 2000. HSP70 stimulates cytokine production through a CD14-dependant pathway, demonstrating its dual role as a chaperone and cytokine. Nat. Med. 6(4):435–42 Tamura Y, Peng P, Kang L, Daou M, Srivastava PK. 1997. Immunotherapy of tumors with autologous tumor-derived heat shock protein preparations. Science 278:117–20 Lehner T, Bergmeier LA, Wang Y, Tao L, Sing M, Spallek R, van der Zee R. 2000. Heat shock proteins generate beta-chemokines which function as innate adjuvants enhancing adaptive immunity. Eur. J. Immunol. 30(2):594–603 Williams SE, Kounnas MZ, Argraves KM, Argraves WS, Strickland DK. 1994. The alpha 2-macroglobulin receptor/low density lipoprotein receptor-related protein and the receptor-associated protein. An overview. Ann. N. Y. Acad. Sci. 737:1– 13

423

71. Binder RJ, Karimeddini D, Srivastava PK. 2001. Adjuvanticity of alpha 2macroglobulin, an independent ligand for the heat shock protein receptor CD91. J. Immunol. 166(8):4968–72 72. Lanzavecchia A. 1985. Antigen-specific interaction between T and B cells. Nature 314(6011):537–39 73. Regnault A, Lankar D, Lacabanne V, Rodriguez A, Thery C, Rescigno M, Saito T, Verbeek S, Bonnerot C, Ricciardi-Castagnoli P, Amigorena S. 1999. Fc gamma receptor-mediated induction of dendritic cell maturation and major histocompatibility complex class I-restricted antigen presentation after immune complex internalization. J. Exp. Med. 189(2):371–80 74. Ohashi K, Burkart V, Flohe S, Kolb H. 2000. Cutting edge: Heat shock protein 60 is a putative endogenous ligand of the toll-like receptor-4 complex. J. Immunol. 164(2):558–61 75. Panjwani NN, Popova L, Febbraio M, Srivastava PK. 2000. The CD36 scavenger receptor as a receptor for gp96. Cell Stress Chaperones 5:391 76. Hilf N, Singh-Jasuja H, Scherer U, Gouttefangeas C, Rammennsee HG, Schild H. 2000. The heat shock protein gp96 binds specifically to human platelets. Cell Stress Chaperones 5:386 77. Botzler C, Li G, Issels RD, Multhoff G. 1998. Definition of extracellular localized epitopes of Hsp70 involved in an NK immune response. Cell Stress Chaperones 3(1):6–11 78. More S, Breloer M, Fleischer B, von Bonin A. 1999. Activation of cytotoxic T cells in vitro by recombinant gp96 fusion proteins irrespective of the ‘fused’ antigenic peptide sequence. Immunol. Lett. 69(2):275–82 79. Berwin B, Reed RC, Nicchitta CV. 2001. Virally induced lytic cell death elicits the release of immunogenic grp94/gp96. J. Biol. Chem. 276(24):21083–88 80. Melcher A, Todryk S, Hardwick N, Ford

4 Feb 2002

11:36

424

81.

Annu. Rev. Immunol. 2002.20:395-425. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

82.

83.

84.

85.

85a.

86.

86a.

AR

AR152-14.tex

AR152-14.SGM

LaTeX2e(2001/05/10)

P1: GKD

SRIVASTAVA M, Jacobson M, Vile RG. 1998. Tumor immunogenicity is determined by the mechanism of cell death via induction of heat shock protein expression. Nat. Med. 4(5):581–87 Gallucci S, Lolkema M, Matzinger P. 1999. Natural adjuvants: endogenous activators of dendritic cells. Nat. Med. 5(11):1249–55 Sauter B, Albert ML, Francisco L, Larsson M, Somersan S, Bhardwaj N. 2000. Consequences of cell death: exposure to necrotic tumor cells, but not primary tissue cells or apoptotic cells, induces the maturation of immunostimulatory dendritic cells. J. Exp. Med. 191(3):423–34 Ishii KJ, Suzuki K, Coban C, Takeshita F, Itoh Y, Matoba H, Kohn LD, Klinman DM. 2001. Genomic DNA released by dying cells induces the maturation of APCs. J. Immunol. 167(5):2602–7 Altmeyer A, Maki RG, Feldweg AM, Heike M, Protopopov VP, Masur SK, Srivastava PK. 1996. Tumor-specific cell surface expression of the KDEL containing, endoplasmic reticular heat shock protein gp96. Int. J. Cancer. 69(4):340–49 Booth C, Koch GL. 1989. Perturbation of cellular calcium induces secretion of luminal ER proteins. Cell 59(4):729–37 Somersan S, Larsson M, Fonteneau JF, Basu S, Srivastava PK, Bhardwaj N. 2001. Primary tumor tissue lysates are enriched in heat shock proteins and induce the maturation of human dendritic cells. J. Immunol. In press Menoret A, Patry Y, Burg C, Le Pendu J. 1995. Co-segregation of tumor immunogenicity with expression of inducible but not constitutive hsp70 in rat colon carcinomas. J. Immunol. 155:740–47 Vanaja DK, Grossmann ME, Celis E, Young CY. 2000. Tumor prevention and antitumor immunity with heat shock protein 70 induced by 15-deoxy-delta12,14prostaglandin J2 in transgenic adenocarcinoma of mouse prostate cells. Cancer Res. 60(17):4714–18

87. Clark PR, Menoret A. 2001. The inducible Hsp70 as a marker of tumor immunogenicity. Cell Stress Chaparones 6(2): 121–25 88. Callahan M, Jacquin C, Chaillot D, Clark P, Menoret A. Chaperoning of immunogenic peptides by Hsp70 and Hsc70 is redox dependent. 11th Int. Congress of Immunol. Stockholm 2001 89. Fadok VA, Bratton DL, Rose DM, Pearson A, Ezekewitz RA, Henson PM. 2000. A receptor for phosphatidylserinespecific clearance of apoptotic cells. Nature 405(6782):85–90 90. Robert J, Menoret A, Basu S, Cohen N, Srivastava PK. 2001. Phylogenetic conservation of the molecular and immunological properties of the chaperones gp96 and hsp70. Eur. J. Immunol. 31(1):186–95 91. Menoret A, Niswonger ML, Altmeyer A, Srivastava PK. 2001. An ER protein implicated in chaperoning peptides to MHC class I is an aminopeptidase. J. Biol. Chem. 276(36):33313–18 92. Srivastava PK. 2000. Immunotherapy of human cancer lessons from mice. Nat. Immunol. 1:363–66 93. Chandawarkar RY, Wagh MS, Srivastava PK. 1999. The dual nature of specific immunological activity of tumor-derived gp96 preparations. J. Exp. Med. 189 (9):1437–42 94. Kovalchin JT, Murthy AS, Horattas MC, Guyton DP, Rajiv Y, Chandawarkar RY. 2001. Determinants of efficacy of immunotherapy with tumor-derived heat shock protein gp96. Cancer Immunity 1:7 95. Lammert E, Stevanovic S, Brunner J, Rammensee HG, Schild H. 1997. Protein disulfide isomerase is the dominant acceptor for peptides translocated into the endoplasmic reticulum. Eur. J. Immunol. 27(7):1685–90 96. Spee P, Neefjes J. 1997. TAP-translocated peptides specifically bind proteins in the endoplasmic reticulum, including gp96, protein disulfide isomerase and calreticulin. Eur. J. Immunol. 27(9):2441–49

4 Feb 2002

11:36

AR

AR152-14.tex

AR152-14.SGM

LaTeX2e(2001/05/10)

P1: GKD

Annu. Rev. Immunol. 2002.20:395-425. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

ROLE OF HSPs IN IMMUNE RESPONSE 97. Binder RJ, Blachere NE, Srivastava PK. 2001. Heat shock protein-chaperoned peptides but not free peptides introduced into the cytosol are presented efficiently by major histocompatibility complex I molecules. J. Biol. Chem. 276(20):17163–71 97a. Li Z, Menoret A, Srivastava PK. 2002. Roles of heat shock proteins in antigen presentation and cross-presentation. Curr. Opin. Immunol. In press 98. Ullrich SJ, Robinson EA, Law LW, Willingham M, Appella E. 1986. A mouse tumor-specific transplantation antigen is a heat shock-related protein. Proc. Natl. Acad. Sci. USA 83:3121–25 99. Wang XY, Kazim L, Repasky EA, Subjeck JR. 2001. Characterization of heat shock protein 110 and glucose-regulated protein 170 as cancer vaccines and the effect of fever-range hyperthermia on vaccine activity. J. Immunol. 166(1):490– 97 100. Palladino MA, Srivastava PK, Oettgen HF, DeLeo AB. 1987. Expression of a shared tumor-specific antigen by two chemically induced BALB/c sarcomas. I. Detection by a cloned cytotoxic T cell line. Cancer Res. 47:5074–79 101. Nicchitta CV. 1998. Biochemical, cell biological and immunological issues

102.

103.

104.

105.

106.

425

surrounding the endoplasmic reticulum chaperone GRP94/gp96. Curr. Opin. Immunol. 10(1):103–9 Janetzki S, Blachere NE, Srivastava PK. 1998. Generation of tumor-specific cytotoxic T lymphocytes and memory T cells by immunization with tumor-derived heat shock protein gp96. J. Immunother. 21(4):269–76 Yedavelli SPK, Guo L, Daou ME, Srivastava PK. 1999. Preventive and therapeutic effect of tumor derived heat shock protein, gp96, in an experimental prostate cancer model. Int. J. Mol. Med. 4:243–48 Graner M, Raymond A, Akporiaye E, Katsanis E. 2000. Tumor-derived multiple chaperone enrichment by free-solution isoelectric focusing yields potent antitumor vaccines. Cancer Immunol. Immunother. 49(9):476–84 Ciupitu AM, Petersson M, O’Donnell CL, et al. 1998. Immunization with a lymphocytic choriomeningitis virus peptide mixed with heat shock protein 70 results in protective antiviral immunity and specific cytotoxic T lymphocytes. J. Exp. Med. 187(5):685–91 Blachere NE, Udono H, Janetzki S, Li Z, Heike M, Srivastava PK. 1993. Heat shock protein vaccines against cancer. J. Immunother. 14(4):352–56

P1: FRK

February 1, 2002

12:47

Annual Reviews

AR152-FM

Annual Review of Immunology Volume 20, 2002

Annu. Rev. Immunol. 2002.20:395-425. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

CONTENTS FRONTISPIECE, Charles A. Janeway, Jr. A TRIP THROUGH MY LIFE WITH AN IMMUNOLOGICAL THEME, Charles A. Janeway, Jr.

xiv 1

THE B7 FAMILY OF LIGANDS AND ITS RECEPTORS: NEW PATHWAYS FOR COSTIMULATION AND INHIBITION OF IMMUNE RESPONSES, Beatriz M. Carreno and Mary Collins

29

MAP KINASES IN THE IMMUNE RESPONSE, Chen Dong, Roger J. Davis, and Richard A. Flavell

PROSPECTS FOR VACCINE PROTECTION AGAINST HIV-1 INFECTION AND AIDS, Norman L. Letvin, Dan H. Barouch, and David C. Montefiori T CELL RESPONSE IN EXPERIMENTAL AUTOIMMUNE ENCEPHALOMYELITIS (EAE): ROLE OF SELF AND CROSS-REACTIVE ANTIGENS IN SHAPING, TUNING, AND REGULATING THE AUTOPATHOGENIC T CELL REPERTOIRE, Vijay K. Kuchroo, Ana C. Anderson, Hanspeter Waldner, Markus Munder, Estelle Bettelli, and Lindsay B. Nicholson

55 73

101

NEUROENDOCRINE REGULATION OF IMMUNITY, Jeanette I. Webster, Leonardo Tonelli, and Esther M. Sternberg

125

MOLECULAR MECHANISM OF CLASS SWITCH RECOMBINATION: LINKAGE WITH SOMATIC HYPERMUTATION, Tasuku Honjo, Kazuo Kinoshita, and Masamichi Muramatsu

165

INNATE IMMUNE RECOGNITION, Charles A. Janeway, Jr. and Ruslan Medzhitov

KIR: DIVERSE, RAPIDLY EVOLVING RECEPTORS OF INNATE AND ADAPTIVE IMMUNITY, Carlos Vilches and Peter Parham ORIGINS AND FUNCTIONS OF B-1 CELLS WITH NOTES ON THE ROLE OF CD5, Robert Berland and Henry H. Wortis E PROTEIN FUNCTION IN LYMPHOCYTE DEVELOPMENT, Melanie W. Quong, William J. Romanow, and Cornelis Murre

197 217 253 301

LYMPHOCYTE-MEDIATED CYTOTOXICITY, John H. Russell and Timothy J. Ley

SIGNAL TRANSDUCTION MEDIATED BY THE T CELL ANTIGEN x

323

P1: FRK

Annu. Rev. Immunol. 2002.20:395-425. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

February 1, 2002

12:47

Annual Reviews

AR152-FM

CONTENTS

xi

RECEPTOR: THE ROLE OF ADAPTER PROTEINS, Lawrence E. Samelson INTERACTION OF HEAT SHOCK PROTEINS WITH PEPTIDES AND ANTIGEN PRESENTING CELLS: CHAPERONING OF THE INNATE AND ADAPTIVE IMMUNE RESPONSES, Pramod Srivastava CHROMATIN STRUCTURE AND GENE REGULATION IN THE IMMUNE SYSTEM, Stephen T. Smale and Amanda G. Fisher PRODUCING NATURE’S GENE-CHIPS: THE GENERATION OF PEPTIDES FOR DISPLAY BY MHC CLASS I MOLECULES, Nilabh Shastri, Susan

371

Schwab, and Thomas Serwold

395 427

463

THE IMMUNOLOGY OF MUCOSAL MODELS OF INFLAMMATION, Warren Strober, Ivan J. Fuss, and Richard S. Blumberg T CELL MEMORY, Jonathan Sprent and Charles D. Surh

GENETIC DISSECTION OF IMMUNITY TO MYCOBACTERIA: THE HUMAN MODEL, Jean-Laurent Casanova and Laurent Abel ANTIGEN PRESENTATION AND T CELL STIMULATION BY DENDRITIC CELLS, Pierre Guermonprez, Jenny Valladeau, Laurence Zitvogel, Clotilde Th´ery, and Sebastian Amigorena

495 551 581

621

NEGATIVE REGULATION OF IMMUNORECEPTOR SIGNALING, Andr´e Veillette, Sylvain Latour, and Dominique Davidson

669

CPG MOTIFS IN BACTERIAL DNA AND THEIR IMMUNE EFFECTS, Arthur M. Krieg

PROTEIN KINASE Cθ

709 IN

T CELL ACTIVATION, Noah Isakov and Amnon

Altman

RANK-L AND RANK: T CELLS, BONE LOSS, AND MAMMALIAN EVOLUTION, Lars E. Theill, William J. Boyle, and Josef M. Penninger PHAGOCYTOSIS OF MICROBES: COMPLEXITY IN ACTION, David M. Underhill and Adrian Ozinsky

761 795 825

STRUCTURE AND FUNCTION OF NATURAL KILLER CELL RECEPTORS: MULTIPLE MOLECULAR SOLUTIONS TO SELF, NONSELF DISCRIMINATION, Kannan Natarajan, Nazzareno Dimasi, Jian Wang, Roy A. Mariuzza, and David H. Margulies

853

INDEXES Subject Index Cumulative Index of Contributing Authors, Volumes 1–20 Cumulative Index of Chapter Titles, Volumes 1–20

ERRATA An online log of corrections to Annual Review of Immunology chapters (if any, 1997 to the present) may be found at http://immunol.annualreviews.org/

887 915 925

8 Feb 2002

9:59

AR

AR152-15.tex

AR152-15.SGM

LaTeX2e(2001/05/10) P1: GSR 10.1146/annurev.immunol.20.100301.064739

Annu. Rev. Immunol. 2002. 20:427–62 DOI: 10.1146/annurev.immunol.20.100301.064739 c 2002 by Annual Reviews. All rights reserved Copyright °

CHROMATIN STRUCTURE AND GENE REGULATION IN THE IMMUNE SYSTEM Stephen T. Smale1 and Amanda G. Fisher2 Annu. Rev. Immunol. 2002.20:427-462. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

1

Howard Hughes Medical Institute and Department of Microbiology, Immunology, and Molecular Genetics, University of California, Los Angeles, California 90095-1662; e-mail: [email protected] 2 Lymphocyte Development Group, MRC Clinical Sciences Centre, Imperial College School of Medicine, Hammersmith Hospital, London W12 ONN UK; e-mail: [email protected]

Key Words nucleosomes, heterochromatin, transcription, interleukin-4 ■ Abstract The development of the immune system and the host response to microbial infection rely on the activation and silencing of numerous, differentially expressed genes. Since the mid-1980s, a primary goal has been to identify transcription factors that regulate specific genes and specific immunological processes. More recently, there has been a growing appreciation of the role of chromatin structure in gene regulation. Before most activators of a gene access their binding sites, a transition from a condensed to a decondensed chromatin structure appears to take place. The activation of transcription is then accompanied by the remodeling of specific nucleosomes. Conversely, the acquisition of a more condensed chromatin structure is often associated with gene silencing. Chromatin structure is a particularly significant contributor to gene regulation because it is likely to be a major determinant of cell identity and cell memory. That is, the propagation of decondensed chromatin at specific loci through DNA replication and cell division helps a cell remember which genes are expressed constitutively in that cell type or are poised for expression upon exposure to a stimulus. Here we review recent progress toward understanding the role of chromatin in the immune system. The interleukin-4 gene serves as a primary model for exploring the events involved in the acquisition and heritable maintenance of a decondensed chromatin structure. Studies of the interleukin-12 p40 and interferon-β genes are then reviewed for insight into the mechanisms by which the remodeling of specific nucleosomes in the vicinity of a promoter can contribute to rapid activation following cell stimulation. Finally, basic principles of gene silencing are discussed.

INTRODUCTION The nucleus has long been known as a complex organelle faced with the enormous challenge of accommodating almost all of a cell’s DNA. The DNA must be packaged and organized in a manner compatible with a number of nuclear events, including differential gene transcription, DNA replication, gene recombination, 0732-0582/02/0407-0427$14.00

427

1 Feb 2002

10:23

Annu. Rev. Immunol. 2002.20:427-462. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

428

AR

SMALE

AR152-15.tex

¥

AR152-15.SGM

LaTeX2e(2001/05/10)

P1: GSR

FISHER

and cell division. The discovery of histones, nucleosomes, and higher order chromatin structures provides a starting point for explaining how these challenges are met (1). In an interphase nucleus, DNA is incorporated into a nucleosome fiber with a diameter of 11 nm. Each nucleosome contains a core histone octamer composed of two molecules of histones H2A, H2B, H3, and H4, and often one linker histone, H1 or H5. The 11-nm fiber is assembled into a higher-order structure known as a 30-nm filament. The 30-nm filament, in turn, can assemble into more condensed, higher-order structures that remain poorly defined. To package the DNA successfully, the bulk of the genome must be assembled into condensed chromatin (1). In a given cell type, decondensed chromatin is primarily associated with genes that must be accessible to the machinery involved in processes such as transcription and recombination. The transitions between condensed and decondensed chromatin structures are thought to be critical for regulating these processes. Decondensation of a locus is likely to be stimulated by cell type- and developmental stage–specific regulatory proteins, although the regulatory proteins directly responsible for decondensation have not been clearly defined for any mammalian gene. Although chromatin decondensation is likely to be a prerequisite to transcriptional activation, nucleosomes within a fully decondensed locus continue to prevent the binding of many transcription factors and the formation of a transcription preinitiation complex at the transcription start site. For occupancy by the full complement of DNA-binding proteins required for transcriptional activation, nucleosomes in the vicinity of a control region must be remodeled (i.e., altered in conformation) or displaced. Nucleosome remodeling and the decondensation of higher-order chromatin structures appear to rely on the activities of two classes of multiprotein complexes: ATP-dependent nucleosome remodeling complexes and histone modification complexes (2–4). ATP-dependent remodeling complexes use the energy of ATP hydrolysis to alter nucleosome conformation (2, 3). Histone modification complexes catalyze the covalent modification of the N-terminal tails of core histones; histones possessing certain modifications appear to form less rigid nucleosome fibers and are recognized by other proteins that contribute to gene regulation, including ATP-dependent remodeling complexes (4). The basic principles of chromatin decondensation and nucleosome remodeling have been the subject of several recent reviews (2–8). In this article, we review, for a general immunology audience, recent progress toward understanding the role of chromatin in the immune system. The focus is on the contributions of chromatin toward the regulation of specific model genes, in particular the interleukin-4 (IL-4), IL-12 p40, and interferon-β (IFN-β) genes. It is perhaps noteworthy that this article does not catalog the many interactions reported between transcription factors and chromatin modifying complexes, in part because the physiological relevance of only a small percentage of these interactions has been confirmed. An additional omission is a discussion of the contributions of chromatin to the recombination and expression of antigen receptor genes. This important topic, which has been studied in considerable depth (9–11), deserves separate consideration.

1 Feb 2002

10:23

AR

AR152-15.tex

AR152-15.SGM

LaTeX2e(2001/05/10)

P1: GSR

CHROMATIN IN THE IMMUNE SYSTEM

429

Annu. Rev. Immunol. 2002.20:427-462. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

CYTOKINE GENE ACTIVATION The interleukin-4 (IL-4) locus serves as an appropriate model for summarizing current knowledge of the contributions of chromatin structure to gene activation in the immune system. IL-4 gene regulation mechanisms have been studied intensively because IL-4 expression helps to define cells of the T helper 2 (Th2) lineage (12, 13). Elucidation of the IL-4 regulatory mechanisms will advance our understanding of the Th1/Th2 lineage decision. Recent reviews have described our accumulated knowledge of the molecular events that contribute to this lineage decision (12, 13). To use the IL-4 gene as a model for introducing the chromatin field to a general immunology audience, we present only a brief overview of the key features of this system and then focus on the role of chromatin in IL-4 gene activation. In na¨ıve T helper precursor (Thp) cells, the IL-4 gene, which characterizes the Th2 lineage, and the IFN-γ gene, which characterizes the Th1 lineage, are silent (12, 13). Upon T cell receptor (TCR) engagement and costimulation, the Thp cells begin to choose between the Th1 and Th2 cell fates (Figure 1). Although the basic steps involved in this lineage decision are not fully understood, current data suggest that the process begins with the rapid activation of both the IFN-γ and IL-4 genes (14). It is not yet clear whether these two genes are initially activated in the same or in different subsets of the stimulated cells. An understanding of this issue is complicated by the fact that each gene is expressed in only a fraction of the cells

Figure 1 Diagram of CD4+ T helper cell differentiation. Adapted from Glimcher & Murphy (12).

1 Feb 2002

10:23

Annu. Rev. Immunol. 2002.20:427-462. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

430

AR

SMALE

AR152-15.tex

¥

AR152-15.SGM

LaTeX2e(2001/05/10)

P1: GSR

FISHER

that acquire competence for expression (15). That is, most of the events required for transcription initiation at the IFN-γ and IL-4 genes may occur in most or all cells, but each gene may actually be transcribed in a small percentage of the cells (see below). Following the antigen stimulation and costimulation of Thp cells, high concentrations of IL-4 mRNA or IFN-γ mRNA have been detected within 1–2 h (14). The initial activation of the IL-4 gene appears to depend on the expression of transcription factor GATA-3 (see below), whereas the initial activation of the IFN-γ gene depends on the expression of T-bet (Figure 1) (12). Following the initial activation events, lineage commitment and expansion of Th1 and Th2 cells requires, in addition to antigen stimulation and costimulation, signaling via the IL-12 and IL-4 receptors, respectively. IL-12 signaling leads to Stat4 activation, which contributes to heritable competence for IFN-γ gene expression and heritable silencing of the IL-4 gene. In this context, the term heritable refers to the cell memory phenomenon, in which a newly developed Th1 cell and virtually all of its progeny remain permanently competent for IFN-γ expression, but incompetent for IL-4 expression. Conversely, IL-4 signaling leads to Stat6 activation, which contributes to heritable competence for IL-4 gene expression and heritable silencing of the IFN-γ gene. Thus, signaling via TCR molecules, costimulatory molecules, and different cytokine receptors is required for Th1 and Th2 lineage commitment and for heritable maintenance of the Th1 and Th2 phenotypes. Lineage commitment is also influenced by other factors, including cell-cycle progression (14, 16, 17). The potential contributions of cell-cycle progression are intriguing but are not discussed here because of space constraints and the complexity of current controversies. Following the initial burst of IL-4 expression during the differentiation of Thp cells into mature Th2 cells, the IL-4 gene becomes transcriptionally inactive (12). However, upon subsequent exposure to an antigen-MHC complex, IL-4 transcription is again rapidly induced. This secondary induction does not require costimulation or IL-4 signaling. The rapid and exclusive expression of Th2 cytokine genes in the absence of Th2-promoting conditions confirms that the cells have maintained competence for expression of these genes and have retained Th1 cytokine genes in a silent state. With respect to chromatin structure, the central goals are to elucidate the mechanisms that allow the IL-4 locus to become permissive for transcription initiation during the primary and secondary activation events and to resist activation upon differentiation toward the Th1 lineage. Of equal interest are the heritable tags that allow mature Th2 cells to selectively activate the IL-4 gene in the absence of costimulation and IL-4 signaling, and that allow mature Th1 cells to maintain the IL-4 locus in a silent state. As discussed below, these heritable tags could include DNA methylation of the IL-4 locus or covalent modification of chromatin components associated with the locus; or the tags could be further removed from the IL-4 locus and include heritable changes in the expression of activators and repressors of IL-4 transcription.

1 Feb 2002

10:23

AR

AR152-15.tex

AR152-15.SGM

LaTeX2e(2001/05/10)

P1: GSR

CHROMATIN IN THE IMMUNE SYSTEM

431

Annu. Rev. Immunol. 2002.20:427-462. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

Coordinate Regulation of a Th2 Cytokine Gene Cluster The human IL-4 locus is positioned on human chromosome 5 (5q31) and the syntenic region of mouse chromosome 11 (18). In all mammalian species that have been examined, the IL-4 gene is contained within a 150-kb chromosomal region that includes two other Th2 cytokine genes, the IL-13 and IL-5 genes (18). Because these three genes are expressed in Th2 cells, broad chromatin alterations may contribute to their coordinate regulation. Many Th2 cells express these genes from only one of the two alleles (15, 19–21). This finding led to the recent demonstration that, within cells that express only one allele of each of the three genes, the frequency with which the expressed alleles were found on the same chromosome was much higher than would be expected if the distribution were random (22). This observation supports the notion that the IL-4, IL-13, and IL-5 genes may be coordinately controlled. One attractive hypothesis to explain the coordinate regulation is that the chromosomal region containing the clustered cytokine genes may be assembled into a heterochromatic structure in Thp cells and in other cell lineages in which the gene is not expressed. Upon Th2 differentiation, the broad region may be coordinately decondensed. Heterochromatin, discussed below in greater depth, is defined cytologically as a region of the nucleus that is stained darkly by DNA dyes, suggesting a high DNA density (5, 6). Within these regions, nucleosomes are thought to be assembled into higher order chromatin structures in which DNA is more compact than within euchromatic regions of the nucleus (7). Euchromatic DNA is stained lightly by DNA dyes and correlates with the locations of actively transcribed genes. Little is known about the structure of heterochromatin, although nucleosomes within heterochromatin appear to be compact, highly ordered, deacetylated, and may contain other modifications, such as histone methylation, that promote assembly into an appropriate, condensed structure (4). Furthermore, CpG dinucleotides and some non-CpG sites appear to be methylated within heterochromatic DNA (8). Recent studies performed in yeast suggest that heterochromatin leads to a transcriptionally silent state by preventing genes from undergoing dynamic changes in chromatin structure (23). One weakness of the hypothesis that the Th2 cytokine locus is assembled into a broad region of heterochromatin in non-expressing cells is that between the IL-13 and IL-5 genes lies the RAD50 gene, a constitutively active gene that encodes a DNA repair enzyme (18). The close juxtaposition of constitutively active (RAD50) and tightly regulated (IL-5 and IL-13) genes highlights our limited knowledge of chromatin structure at complex loci. Does this chromosomal region contain two broad domains of condensed chromatin separated by a domain of decondensed chromatin containing the RAD50 gene? Is each gene regulated by an independent chromatin domain, with a common enhancer region responsible for the coordinate regulation? If the locus is heterochromatic on the basis of cytological criteria, do unique properties of the RAD50 control regions allow it to be transcribed from a heterochromatic location, as has been observed for a small subset of genes in

1 Feb 2002

10:23

Annu. Rev. Immunol. 2002.20:427-462. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

432

AR

SMALE

AR152-15.tex

¥

AR152-15.SGM

LaTeX2e(2001/05/10)

P1: GSR

FISHER

lower eukaryotes? Although these and other questions remain to be answered, immunoFISH experiments have recently demonstrated that the silent IL-4 locus in Th1 cells is indeed positioned in the vicinity of cytological heterochromatin (14), consistent with the basic hypothesis that heterochromatin contributes to IL-4 gene regulation. One final piece of evidence that the cytokine genes are coordinately regulated emerged from an analysis of mice containing a yeast artificial chromosome (YAC) transgene encompassing the entire Th2 cytokine cluster (18, 24). Deletion of a conserved 400-bp DNA region located between the IL-4 and IL-13 genes (the CNS-1 or HSS1-2 region, see below) diminished expression of all three cytokines but had no effect on expression of the RAD50 gene (18). This modest reduction in expression of the three cytokines was recently confirmed by targeted deletion of the CNS-1 region within the endogenous locus (25). These results demonstrate that all three cytokine genes are regulated to some extent by at least one common control region. It is intriguing that the reduced expression of the IL-4 gene following deletion of the CNS-1 region was due, at least in part, to a reduction in the percentage of IL-4-expressing cells within a population (18, 25). One possibility is that the CNS-1 region increases the probability that the chromatin associated with each allele will form a stable, decondensed structure that is competent for activation. That is, the control region could contribute to the coordinate decondensation of the cytokine cluster. Another possibility, however, is that the control region acts downstream of the chromatin-related events by increasing the probability that a transcription preinitiation complex will form at the cytokine promoters. This possibility is consistent with previous in vivo and in vitro studies, which showed that enhancers and transcriptional activators can influence the number of alleles containing productive preinitiation complexes in a chromatin-independent manner (26, 27). To distinguish between these possibilities, it will be important to determine whether the absence of IL-4 expression in individual cells is heritable. That is, do non-expressing clones remain stably incompetent for expression when they are propagated? Further evidence that the control region acts at the level of chromatin structure could be obtained by determining whether the fraction of cells that fail to express the gene exhibit a more condensed chromatin structure than do the expressing cells, as determined by DNase I sensitivity or DNA methylation analyses (see below). It is noteworthy that a careful examination of the IL-4, IL-5, and IL-13 genes revealed that, although they are often expressed in the same cells, it was not uncommon to identify cell clones that expressed only one or two of these genes (22). Furthermore, it was not uncommon to find cells expressing both alleles of one gene and only one allele of the others (22). Therefore, although gene activation may be accompanied by the coordinate decondensation of the locus from a heterochromatic to a euchromatic state, mechanisms must exist that allow each gene to be regulated independently.

1 Feb 2002

10:23

AR

AR152-15.tex

AR152-15.SGM

LaTeX2e(2001/05/10)

P1: GSR

Annu. Rev. Immunol. 2002.20:427-462. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

CHROMATIN IN THE IMMUNE SYSTEM

433

To explore the general mechanism of coordinate regulation in greater depth, an appropriate starting point would be a careful examination of the DNase I sensitivity and histone modification status of the entire locus, including the cytokine genes, RAD50 gene, and the flanking regions. DNase I sensitivity as a means of monitoring global chromatin structure was first described in 1976 by Weintraub & Groudine (28) in conjunction with studies of the chicken globin locus. To perform this assay, aliquots of isolated cell nuclei are treated with increasing concentrations of pancreatic DNase I. Genomic DNA is then isolated and cleaved with a restriction enzyme. The integrity of the restriction fragment of interest is then monitored by Southern blot. Using a modified version of this strategy, restriction fragments within the globin locus were found to be digested with much lower concentrations of DNase I in nuclei from erythrocytes than in nuclei from nonerythroid cells. The region of DNase I sensitivity spanned the entire globin locus, suggesting that the entire locus was assembled into a decondensed chromatin structure. It is tempting to speculate that the transition from DNase I resistance to sensitivity corresponds to a transition from heterochromatin to euchromatim. However, the formal relationship between DNase I resistance and cytological heterochromatin has not been confirmed. In the case of the Th2 cytokine locus, it would be interesting to determine whether greater sensitivity to DNase I digestion is observed in Th2 cells than in na¨ıve T cells or Th1 cells, and whether the sensitivities of the four genes are coordinately regulated. Is the RAD50 gene sensitive only when the Th2 genes are sensitive, or is this gene constitutively sensitive, which would suggest a constitutively decondensed chromatin structure? Furthermore, in a cell line expressing only one Th2 gene, is the entire locus or only the expressed gene sensitive to DNase I digestion? These results may provide insight into the degree to which chromatin structure is coordinately regulated. The experiments are likely to be complicated, however, by the existence of monoallelic expression; to obtain meaningful results, it may be necessary to use probes that are allele-specific. Another property that may provide insight into the mechanism of coordinate regulation is the modification status of nucleosomes that span the locus. It is now well-established that core histones are subject to a range of posttranslational modifications, mostly at their N-terminal tails located at the nucleosome surface (4). Acetylation is the most widely studied histone modification, although several other posttranslational modifications have been characterized, including methylation and phosphorylation. Acetylation correlates with increased transcriptional potential and is thought to promote the formation of a nucleosomal structure that is accessible to the transcription machinery. Histone acetylation can be restricted to one or two nucleosomes in the vicinity of a promoter, but it can also extend throughout an entire locus. At the chicken and mouse β-globin loci, acetylation spans the locus and correlates with DNase I sensitivity (29, 30). Thus, an analysis of histone modifications throughout the Th2 cytokine locus could complement the DNase I sensitivity studies and help to answer the questions posed above.

1 Feb 2002

10:23

434

AR

SMALE

AR152-15.tex

¥

AR152-15.SGM

LaTeX2e(2001/05/10)

P1: GSR

FISHER

Annu. Rev. Immunol. 2002.20:427-462. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

Chromatin Changes during Th2 Differentiation: Establishing Competence for IL-4 Gene Expression Although the mechanisms by which Th2 cytokine genes are coordinately regulated remain poorly understood, much more is known about the IL-4 gene itself. Two biochemical properties of the IL-4 locus, DNase I hypersensitivity and DNA methylation, suggest that the chromatin structure is substantially altered during Th2 differentiation and gene activation. The DNase I hypersensitivity assay provides very different information about chromatin structure than does the DNase I sensitivity assay (31). The two assays are performed using a similar procedure, in which aliquots of nuclei are treated with increasing concentrations of DNase I, followed by restriction enzyme cleavage of purified genomic DNA and Southern blot analysis. In the DNase I hypersensitivity assay, the goal is to identify specific sites that are cleaved with unusually low concentrations of DNase I. In contrast, the DNase I sensitivity assay monitors the overall efficiency of digestion of a broad DNA region. Although a decondensed locus is generally more sensitive to DNase I digestion than a condensed locus, the decondensed locus is still assembled into nucleosomal structures that limit DNase I digestion. Hypersensitivity to DNase I digestion is often observed at specific sites within the locus that lack nucleosomes or contain nucleosomes with disrupted structures due to the binding of transcription factors. DNase I hypersensitive sites were first described in the heat-shock genes of the fruitfly Drosophila melanogaster (32) and were hypothesized to correspond to transcriptional control regions. Numerous studies have confirmed that DNase I hypersensitive sites often (but not always) correspond to promoters, enhancers, and other types of control regions, including locus control regions (LCRs), silencers, insulators, and matrix attachment regions (31). In the IL-4 locus, two DNase I hypersensitive sites were observed that are present in all T helper cells (Thp, Th1, and Th2) (Figure 2a). One of these sites, HSS3, is located 9.5 kb upstream of the transcription start site for the IL-4 gene (33). The second site, site IV, is located approximately 1 kb downstream of the final exon of the IL-4 gene (34). The functional significance of these sites is not well established, although site IV has been reported to possess silencer activity (35). It is not clear whether these sites are bound by sequence-specific DNA binding proteins that contribute to activation after other important control regions are occupied, are bound by proteins that help to maintain the inactive state of the locus (consistent with the silencer results), or simply correspond to functionally irrelevant regions of the locus where the chromatin structure diverges from a standard nucleosomal array, perhaps due to an unusual DNA sequence. Several additional DNase I hypersensitive sites have been described that are observed only in differentiating and mature Th2 cells (Figure 2a). Two of these sites, HSS2 and HSS1 (together referred to as the CNS-1 region), are located 8.5 kb and 8.3 kb upstream of the transcription start site, respectively (33). One, site I, is at the promoter (34). Two others, sites II and III, are located in the second

1 Feb 2002

10:23

AR

AR152-15.tex

AR152-15.SGM

LaTeX2e(2001/05/10)

P1: GSR

Annu. Rev. Immunol. 2002.20:427-462. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

CHROMATIN IN THE IMMUNE SYSTEM

435

Figure 2 DNase I hypersensitive sites within the Th2 cytokine cluster (a) and the IL-4 mini-locus transgene (b). Adapted from Lee et al. (40).

intron, and a sixth, site V, is located approximately 5 kb downstream of the final exon (34). These sites are not observed in Thp or Th1 cells. Sites II, III, and V have been characterized most extensively and appear shortly after differentiation of Thp cells toward the Th2 lineage (34). They therefore appear to correspond to control regions that become occupied by transcription factors only when Thp cells are induced to differentiate into Th2 cells. The absence of these hypersensitive sites in Thp and Th1 cells suggests that the putative control regions are not occupied and are instead assembled into a regular nucleosomal array. It is not clear whether, in Th2 cells, the nucleosomes at these sites have been displaced or have simply undergone a remodeling event. Most interestingly, the DNase I hypersensitive sites are retained in Th2 cells even when the IL-4 gene is not being actively transcribed (34). This result suggests that transcription factor occupancy of these putative control regions is insufficient for transcription. More importantly, the sustained occupancy of these control

1 Feb 2002

10:23

Annu. Rev. Immunol. 2002.20:427-462. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

436

AR

SMALE

AR152-15.tex

¥

AR152-15.SGM

LaTeX2e(2001/05/10)

P1: GSR

FISHER

regions may be critical for the subsequent activation of IL-4 transcription upon TCR engagement in the absence of costimulation or IL-4 signaling. Thus, the factors bound to these control regions are likely to contribute to the memory that is a distinguishing feature of mature Th2 cells. It is noteworthy that this phenomenon, in which a locus is heritably poised for expression even when it is transcriptionally inactive, may be a unique feature of genes that require rapid transcriptional induction in response to a stimulus. By contrast, occupancy of distant control regions and transcriptional activity may be more tightly linked at genes that do not require rapid induction. The DNase I hypersensitivity studies suggest that nucleosomes associated with the hypersensitive sites are displaced or remodeled upon stimulation of Thp cells. However, it is noteworthy that these results provide little information about the global chromatin structure at the IL-4 locus. When the DNase I hypersensitive sites appear, the entire IL-4 locus must be sufficiently decondensed to allow for occupancy of the control regions. However, it is not clear whether these hypersensitive sites appear when the locus is initially decondensed, or whether decondensation occurs independently and at an earlier stage of development, perhaps during the development of Thp cells. One key property of the IL-4 locus that supports the former hypothesis is its DNA methylation status. The occurrence of methylated CpG dinucleotides at the IL-4 locus was high in Thp cells and nonlymphoid cells and was greatly reduced upon Th2 differentiation (16, 34). The reduced methylation that accompanies differentiation suggests a global change in chromatin structure that may correspond to the initial decondensation of the locus. DNA methylation status has long been correlated with gene activity, with a high content of methylated CpG dinucleotides generally observed at transcriptionally silent genes. However, it was not known whether methylation status plays an active role in gene regulation or whether methylation is simply a passive consequence of transcriptional inactivity. Recently, a number of results have provided strong evidence that DNA methylation is indeed an active regulator of chromatin structure (8). One key finding was that methylated nucleotides bind factors that associate with histone deacetylase complexes. The current dogma is that DNA methylation can stimulate the formation of a relatively inaccessible chromatin structure, whereas unmethylated DNA contributes to an accessible structure. It will be particularly important to determine how occupancy of the DNase I hypersensitive regions by transcription factors and the unmethylated state of the DNA contribute to the heritable competence for IL-4 expression in mature Th2 cells. Does the simple expression of lineage-restricted DNA-binding proteins in Th2 cells allow the control regions to be heritably occupied, thereby keeping the gene poised for rapid activation during a subsequent stimulation? In this model, the transcription factors would most likely be displaced from the control regions during each round of DNA replication and cell division, following by repeated, de novo occupancy of the control regions based solely on the continual expression of Th2-restricted factors. Alternatively, is the unmethylated state of the DNA, or a

1 Feb 2002

10:23

AR

AR152-15.tex

AR152-15.SGM

LaTeX2e(2001/05/10)

P1: GSR

Annu. Rev. Immunol. 2002.20:427-462. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

CHROMATIN IN THE IMMUNE SYSTEM

437

particular histone modification pattern, critical for tagging the locus and enabling the transcription factors to reassociate with the control regions after each round of DNA replication and cell division? In this second model, the heritable tags (or imprints) would be placed directly on the IL-4 locus, whereas in the first model, the heritable tags would be on the genes encoding critical regulators of IL-4 transcription (e.g., the GATA-3 gene). Importantly, these two models are not mutually exclusive. To focus momentarily on the concept of heritable tags, it is noteworthy that DNA methylation and histone modification patterns are both attractive candidates for carrying out this function. Unmethylated DNA is not recognized by DNMT1, the DNA methyltransferase that maintains the methylated state of inactive genes by selectively targeting hemimethylated DNA following DNA replication (8). Thus, the unmethylated state can easily be propagated, helping to propagate an accessible chromatin structure. It may also be possible to propagate a histone modification pattern through S phase (36). One hypothesis is that an existing histone modification pattern at a locus could restrict DNA replication to a particular time within S phase, during which nucleosomes possessing the desired modification pattern are exclusively assembled. Alternatively, a histone modification pattern could lead to replication at a specific location within the nucleus. The identification of heritable tags and an understanding of their propagation mechanisms are the primary goals of the epigenetics field. This topic is discussed in slightly more detail in the final section of the article. As described above, the appearance of the DNase I hypersensitive sites and the reduced DNA methylation may be important for IL-4 gene transcription. However, these events are almost certainly insufficient for transcription. This conclusion is based on the fact that the hypersensitive sites are present in mature Th2 cells prior to stimulation and gene activation. In addition, the locus is unmethylated in nonexpressing mature Th2 cells. These observations suggest that additional control regions are required for gene activation. Alternatively, the control regions that are hypersensitive in non-expressing cells may be occupied by only a subset of the factors that are required for transcription; additional factors may bind these regions during activation, or the bound factors may be modified, leading to recruitment of coactivators and the general transcription machinery. Although these possibilities are not mutually exclusive, the first hypothesis is supported by the discovery of an additional hypersensitive site, Va, which appears only when the gene is actively transcribed (Figure 2a) (37). The ability of this region to function as an inducible enhancer in transfection experiments supports its potential role as a key regulator of IL-4 gene transcription (37). The relationship between the Va hypersensitive site and the others remains unknown. The other hypersensitive sites may lead to the formation of a relatively accessible chromatin environment, which may be required for transcription factors to bind the Va site. After the Va region becomes occupied, it may help stimulate transcription in concert with control regions associated with the other DNase I hypersensitive site. Alternatively, the Va region and the promoter may

1 Feb 2002

10:23

438

AR

SMALE

AR152-15.tex

¥

AR152-15.SGM

LaTeX2e(2001/05/10)

P1: GSR

FISHER

be the primary stimulators of transcription, with the other distant control regions responsible primarily for general accessibility of the locus. The former hypothesis is supported by the functional studies described below.

Annu. Rev. Immunol. 2002.20:427-462. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

Functions of Individual Control Regions in the IL-4 Locus The above studies focused primarily on physical properties of the IL-4 locus. The results provide important insight into the molecular events that contribute to gene activation and cell fate. However, detailed functional studies of individual control regions will be required for a full understanding of their precise roles. Most functional studies of IL-4 gene regulation have focused on the promoter (12). The isolated promoter was found to be sufficient for cell type–restricted transcription in transfection assays and transgenic mice (38, 39). However, the efficiency of transcription in the transgenic mice was well below that observed with the endogenous gene. Furthermore, the IL-4 promoter-reporter transgenes yielded highly variable transcriptional activities (39). These findings provided the first functional evidence that distant control regions are required for proper regulation of the endogenous IL-4 gene. The activities of distant control regions have been analyzed most extensively in one recent study (40). This study analyzed, in transgenic mice, the activities of the DNA regions encompassing most of the DNase I hypersensitive sites described above. An additional region that is highly conserved through evolution was also analyzed. For this analysis, the putative control regions were fused individually or in various combinations to the IL-4 promoter and a luciferase reporter gene. After transgenic mice were generated with each construct, luciferase activities in both Th1 and Th2 cells were determined for several different founders. The results obtained were interesting, yet complex. The only DNA fragment that, by itself, acted as a strong transcriptional enhancer contained three of the DNase I hypersensitive sites, HSS1-3. As described above, two of these hypersensitive sites encompass the CNS-1 region, which, when deleted in the context of either a YAC transgene or an endogenous allele, resulted in a modest reduction in expression of the IL-4 gene and the other Th2 cytokine genes (18, 25). This deletion had little effect on the Th2-specificity of IL-4 expression. Consistent with these results, the HSS1-3 region was equally active in Th1 and Th2 cells in the transgenic mice generated by Lee et al. (40). Thus, the region is, by definition, a potent enhancer, yet its function within the intact locus is to moderately boost Th2 cytokine expression. It is interesting to recall that this is the longest and most highly conserved region within the IL-4 locus (18), yet it does not appear to be essential for Th2 specificity. The moderately reduced expression following deletion of CNS-1 from the endogenous locus, relative to the strong enhancer activity of the slightly larger HSS1-3 region in the context of the luciferase reporter transgene, makes it difficult to predict why this region evolved and how it globally influences Th2 cytokine expression. Is its function simply to enhance Th2 cytokine expression by a few-fold?

1 Feb 2002

10:23

AR

AR152-15.tex

AR152-15.SGM

LaTeX2e(2001/05/10)

P1: GSR

Annu. Rev. Immunol. 2002.20:427-462. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

CHROMATIN IN THE IMMUNE SYSTEM

439

Considering the dramatic effect of the CNS-1 deletion on the immune response to important pathogens, the precise control of cytokine expression levels is clearly very important and could be sufficient to explain the evolution of this region (25). Alternatively, is it critical for an additional regulatory event that has not yet been uncovered? Finally, is it partially redundant with another control region, which would suggest that it evolved because of the inherent benefits of redundancy within the genome (i.e., to minimize the impact of fortuitous mutations)? As described earlier, the effect of the CNS-1 deletion on expression efficiency could be due to a role in establishing an accessible chromatin structure or a role in increasing the percentage of alleles that contain productive transcription preinitiation complexes. These different roles are not mutually exclusive. One result which suggests that the region functions, at least in part, at the level of chromatin, is that it was incapable of acting as an enhancer in a transient transfection assay (40). Because transiently transfected plasmids generally do not assemble into physiological chromatin structures (41), this result suggests that the activity of the control region may depend on a chromatin environment. Alternatively, the high plasmid copy number obtained in transiently transfected cells could overwhelm the available transcription factors, preventing the control regions from associating with the full complement of factors required for activity (31). None of the other individual regions analyzed by Lee et al. (40) supported strong enhancer activity. Furthermore, preferential activity in Th2 cells was observed only with the intronic region containing hypersensitive II, which weakly stimulated transcription. Most interestingly, strong, Th2-specific expression was observed only when all of the elements were combined into a mini-locus (Figure 2b). These findings reveal the complexity of the interplay between the factors bound to the various control regions. It is interesting to note that the reporter activity observed in Th2 cells of the transgenic mice containing the minilocus was comparable to that observed in both Th1 and Th2 cells of mice containing only the HSS1-3 region. Thus, the primary effect of the other control regions, when combined, is to suppress IL-4 expression in Th1 cells rather than to enhance expression in Th2 cells. The complexity of these results makes it difficult to predict the potential relevance of each control region for decondensation of the repressed chromatin during Th2 development or for chromatin remodeling events that are more intimately linked to gene transcription. An understanding of these issues is likely to require the targeted deletion of individual control regions, followed by an analysis of the effects on gene transcription and chromatin structure, as determined by analysis of DNase I and restriction enzyme sensitivity, DNA methylation, and histone modifications. One final question is whether additional control regions may exist that were not revealed by the DNase I hypersensitivity studies or the analysis of conserved sequences. This possibility is worth mentioning because the minilocus transgene was highly susceptible to integration-site effects. That is, luciferase activity varied considerably among the mini-locus founders (40). The two types of control regions that facilitate efficient, integration site–independent expression are locus control

1 Feb 2002

10:23

Annu. Rev. Immunol. 2002.20:427-462. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

440

AR

SMALE

AR152-15.tex

¥

AR152-15.SGM

LaTeX2e(2001/05/10)

P1: GSR

FISHER

regions (LCRs) and insulators (42, 43). LCRs, first described in the β-globin locus, exhibit properties often comparable to those of enhancers. However, they possess a greater capacity to alter long-range chromatin structures and to limit the spread of a surrounding repressive chromatin environment. An insulator is a boundary element possessing a specific function. The first boundary element was identified in Drosophila as a DNA region that served as a boundary between decondensed chromatin associated with an active gene and condensed chromatin associated with flanking regions (44). When placed between a chromosomally integrated enhancer and promoter, this boundary element prevented the enhancer from stimulating promoter-driven transcription. Because enhancer activity in this context is thought to depend on its ability to alter chromatin structure, the boundary element’s apparent function was to insulate the promoter from the surrounding chromatin environment. Although LCRs and insulators have been identified at only a handful of loci, one would expect that many loci would possess one or both types of control regions. Thus, the variable activities of the minilocus transgenes suggest that additional IL-4 control regions remain to be identified.

Functions of Individual Transcription Factors at the IL-4 Locus Most control regions that have been characterized span 100–400 bp and contain recognition sites for several DNA-binding proteins (45). Each of the DNase I hypersensitive sites described above is likely to correspond to a bona fide control region. Some of the DNA-binding proteins that occupy the control regions are likely to be induced during Th2 differentiation or cell activation, whereas others are likely to be regulated less tightly. To bind the native control region, these unregulated proteins probably require a decondensed chromatin structure or cooperative interactions with the regulated DNA-binding proteins. The requirement for multiple transcription factors expressed in different, partially overlapping sets of cell types is a basic feature of combinatorial gene regulation strategies (45). A number of DNA-binding proteins have been identified that can bind the IL-4 promoter and distant control regions in T cells (12). Included on this list are two factors that are expressed in Th2 cells, but not in Th1 cells: GATA-3 (46) and c-Maf (47). Several other factors that are expressed more broadly, including NFAT and AP-1, are also capable of binding IL-4 control regions (12). Another candidate for a direct activator of IL-4 gene transcription is Stat-6, although relevant binding sites for this factor have not been reported. Finally, a Th2-restricted, non-DNA-binding factor, NIP45, appears to be critical for efficient IL-4 expression (48). NIP45 may function as a coactivator that facilitates transcriptional activation by interacting with the DNA-binding proteins. The precise, biochemical functions of each of these factors during Th2 differentiation and the induction of IL-4 transcription are not known. It is intriguing that many of these factors and some of the control regions described above, including the CNS-1 region, do not appear to contribute to IL-4 gene induction in mast cells (25, 49). Brown and colleagues (49) have identified several other factors that are required for IL-4 expression in mast cells and have

1 Feb 2002

10:23

AR

AR152-15.tex

AR152-15.SGM

LaTeX2e(2001/05/10)

P1: GSR

Annu. Rev. Immunol. 2002.20:427-462. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

CHROMATIN IN THE IMMUNE SYSTEM

441

proposed models that may provide biological and mechanistic explanations for the dramatic differences. The most intriguing factor required for IL-4 expression in Th2 cells is GATA-3. This factor is expressed in Thp cells and in differentiating and mature Th2 cells (46). However, following stimulation of Thp cells under Th1-promoting conditions, its expression is gradually reduced. Antisense experiments established that GATA-3 is essential for the expression of Th2 cytokines, including IL-4 (46). Furthermore, overexpression of GATA-3 in Th1 cells was sufficient for inducing Th2 cytokine gene expression (46). The induction of Th2 cytokine expression and Th2 differentiation by GATA-3 was found to be independent of IL-4 signaling, which suggests that GATA-3 may be the earliest inducer of these events (50–52). The identification of GATA-3 binding sites within various IL-4 control regions suggests that GATA-3 may induce transcription via direct binding to these sites. In addition, the observation that GATA-3 overexpression can generate most of the DNase I hypersensitive sites within the IL-4 locus suggests that it may be responsible for decondensation of the locus (50–52). In the simplest and most direct scenario, GATA-3 may be capable of accessing its target sites within the IL-4 control regions when they are assembled into a condensed chromatin structure. Upon binding, GATA-3 may recruit histone modification complexes and/or ATP-dependent nucleosome remodeling complexes, leading to chromatin decondensation. Other transcription factors that pre-exist in Thp cells or that are induced upon GATA-3 expression, such as c-Maf, may then bind the decondensed locus. Full occupancy of these control regions would result in the appearance of the DNase I hypersensitive sites and gene transcription. Although the above model is intriguing, very little is known about GATA3’s mechanism of action. One model at the opposite end of the spectrum is that GATA-3 may primarily be an essential activator of genes encoding one or more critical transcription factors that have not yet been discovered. These factors may then bind to sites within the IL-4 locus and promote the decondensation of the locus. Once the locus has been decondensed, GATA-3 and other factors may bind and stimulate transcription. In this model, GATA-3 may contribute to the local remodeling of nucleosomes within the enhancers or promoter, or it may communicate directly or indirectly with the general transcription machinery positioned at the promoter. However, it would not be the direct inducer of chromatin decondensation. Several other models between these two extremes can be imagined. To emphasize the current uncertainties, it is noteworthy that the ability of overexpressed GATA-3 to promote Th2 differentiation and IL-4 expression is independent of its DNA-binding domain (51). Furthermore, Thp cells express GATA-3, yet are clearly less competent for induction of Th2 cytokine expression than cells containing overexpressed GATA-3. GATA-3’s DNA-binding domain may be expendable because the protein may be incorporated into a multiprotein complex containing other DNA-binding proteins, such that its own DNA-binding domain is not essential for binding to control regions when it is overexpressed. Moreover, physiological GATA-3 concentrations may be insufficient for Th2 cytokine

1 Feb 2002

10:23

Annu. Rev. Immunol. 2002.20:427-462. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

442

AR

SMALE

AR152-15.tex

¥

AR152-15.SGM

LaTeX2e(2001/05/10)

P1: GSR

FISHER

expression in Thp cells because of competition with other transcription factors that are expressed in Thp cells or are rapidly induced upon stimulation, such as T-bet (53, 54). Nevertheless, despite these potential explanations for the above results, the possibility will need to be considered that other factors are equally important for Th2 differentiation when GATA-3 is present at physiological concentrations. Although members of the GATA family, like many transcription factors, are capable of interacting with multiprotein complexes implicated in nucleosome remodeling (55), rigorous functional studies will be required to determine whether such interactions involving GATA-3 contribute directly to the decondensation of the IL-4 locus and IL-4 gene transcription. To date, it has been very difficult in mammalian systems to establish the precise functions of specific transcription factors during the activation of an endogenous gene (31). To achieve this goal, a multifaceted approach will almost certainly be required. One goal will be to identify the GATA-3 domains (and amino acids within those domains) that are critical for Th2 differentiation and IL-4 gene induction, ideally by reconstituting GATA3-deficient Thp cells with mutant GATA-3 proteins expressed at physiological concentrations. The proteins and protein complexes that interact with those domains can then be determined. Since multiple factors will be capable of interacting with any given domain in artificial assays, it will be important to establish a close correlation between the amino acids required for the protein-protein interaction and the amino acids required for IL-4 induction and Th2 differentiation. Complementary studies will be necessary to define the cis-acting sequences within the control regions that are essential for nucleosome alterations and gene activation. Although GATA-3 has been studied most extensively, intriguing results with respect to chromatin structure have been obtained for two other regulators of IL-4 transcription, c-Maf and NFAT1. c-Maf is a Th2-restricted factor that, on the basis of gene disruption studies, is known to be critical for IL-4 expression (47, 56). GATA-3 overexpression induces c-Maf expression, suggesting that c-Maf acts downstream of GATA-3 (46). This hypothesis was supported by an elegant experiment performed by Agarwal et al. (37), which revealed that the DNase I hypersensitive sites associated with the IL-4 locus are readily detected in Th2 cells from c-Maf−/− mice. Although c-Maf binds to the IL-4 promoter and possibly to other IL-4 control regions, this result demonstrated that it is not required for decondensation of the locus or for the appearance of the DNase I hypersensitive sites. This result is highly significant because it helps to establish the function of c-Maf using an experiment that does not rely on GATA-3 overexpression. A likely scenario is that c-Maf can bind to the promoter only after the locus has been decondensed. Once bound, it may communicate with coactivators and the general transcription machinery in synergy with other DNA-binding proteins associated with the promoter, leading to transcription initiation. It may also contribute to the local remodeling of nucleosomes in the vicinity of the promoter. NFAT proteins are equally important for IL-4 transcription, yet they are expressed in both Th1 and Th2 cells (12). In Th1 cells, NFAT proteins appear to contribute to the induction of the IFN-γ gene and several other genes (12). Chromatin

1 Feb 2002

10:23

AR

AR152-15.tex

AR152-15.SGM

LaTeX2e(2001/05/10)

P1: GSR

Annu. Rev. Immunol. 2002.20:427-462. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

CHROMATIN IN THE IMMUNE SYSTEM

443

immunoprecipitation (ChIP) experiments provided insight into the mechanism by which NFAT proteins contribute to transcription of different genes in the two different lineages (37). The ChIP results revealed that NFAT1 binds the IL-4 promoter and the enhancer encompassing the Va hypersensitive site in Th2 cells, but not in Th1 cells. In contrast, this same protein binds the IFN-γ promoter in Th1 cells, but not in Th2 cells. The implication of these results is that NFAT1 binding can occur only after a locus has been decondensed. Cooperative binding with Th2-specific factors, such as GATA-3 and c-Maf, could also contribute to Th2-specific binding to the IL-4 locus. Once bound, NFAT1, like c-Maf, may help promote transcription initiation through interactions with coactivators and the general transcription machinery, or by helping to alter the local chromatin structure.

Monoallelic Expression and Allelic Bias One of the most fascinating discoveries to emerge from studies of cytokine gene regulation is the frequent activation of transcription from only one of the two endogenous alleles. Monoallelic expression (or allelic bias) of a cytokine gene was first documented for the IL-2 gene (57), a few years after the discovery that natural killer cell receptor genes are expressed monoallelically (58). Allelic bias of IL-4 gene expression was established by three different groups (19–21) and was recently extended to other Th2 cytokine genes (22). First, Bix & Locksley (19) analyzed F1 heterozygotes of BALB/c mice and Mus musculus casteneus, which contain numerous DNA polymorphisms that allow the two IL-4 alleles to be readily distinguished. In Th2-primed cells and Th2 clones, IL-4 expression was often limited to one allele. In a second study, Riviere et al. (20) used a knockin strategy to introduce a human CD2 gene into the IL-4 locus. Th2 cells often contained transcripts and protein from only one allele, either the wild-type IL-4 allele or the IL-4/hCD2 chimeric allele. Hu-Li et al. (21) obtained similar results after inserting a green fluorescent protein (GFP) gene cassette into one IL-4 allele. The biological significance of monoallelic expression remains unknown and is not discussed here because of the mechanistic focus and space constraints. However, a number of intriguing hypotheses have been presented (15, 19–22). One central property of the allelic bias that is likely to be critical for a mechanistic understanding is that the bias was not heritable during the initial differentiation of Th2 cells (21). Rather, heritability was acquired only after repeated stimulation of the cells. That is, when cells that expressed only the IL-4/GFP chimeric allele during the primary stimulation were isolated and restimulated, no significant bias toward that same allele was observed. In contrast, after multiple stimulations, cells that expressed only the IL-4/GFP chimeric allele were likely to express the same allele during the next stimulation. The mechanisms responsible for allelic bias are not known. However, the transition from a nonheritable to a heritable state suggests that the mechanism underlying the nonheritable bias is likely to differ from the mechanism underlying the heritable bias. Changes in chromatin structure are not necessarily involved in the

1 Feb 2002

10:23

Annu. Rev. Immunol. 2002.20:427-462. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

444

AR

SMALE

AR152-15.tex

¥

AR152-15.SGM

LaTeX2e(2001/05/10)

P1: GSR

FISHER

nonheritable bias. Rather, two general models can be considered. One model is that, in most cells, only one allele may acquire the full complement of transcription factors required for efficient activation. The inefficient acquisition of transcription factors could ultimately be due to insufficient concentrations of the factors. This insufficiency would presumably be linked to unfavorable thermodynamics; for example, the transcriptional activators may not compete effectively with the repressive chromatin environment that exists in the unstimulated Thp cells. The second model, mentioned earlier, is that both alleles may acquire the full complement of transcription factors, but recruitment of the general transcription machinery and/or formation of a productive preinitiation complex may be inefficient (26, 27). It would be difficult to account for the subsequent heritable bias using the above models. Therefore, a different mechanism must be evoked. Most likely, the active or inactive allele acquires a tag that can propagate the active or inactive state. DNA methylation or a specific histone modification pattern are currently the leading candidates for heritable tags (4, 8). The basis of the transition from the nonheritable to heritable state will be particularly interesting to uncover. One of several possible models is that, during the initial stimulation of Thp cells, all alleles acquire properties of fully open chromatin (e.g., unmethylated CpG dinucleotides and acetylated histones). During the initial stimulation and each subsequent stimulation, only one allele is activated in many of the cells because of inefficient occupancy or inefficient preinitiation complex formation, as described above. As the stimulated cells proliferate, any allele that is inactive has the potential to acquire, with a low but significant probability, methylated CpG dinucleotides or histone modifications that promote the formation of a repressive chromatin environment. In other words, with each stimulation, the percentage would increase of cells in which one of the alleles possesses a tag that results in a heritably silent state. Although these and other models can be envisioned on the basis of current knowledge, one additional observation of Hu-Li et al. (21) is difficult to explain: namely, the observation that the extent of the heritable allelic bias varies from clone to clone. That is, one clone may exhibit a weak bias toward one allele (only slightly more protein derived from one allele than the other), whereas another clone may exhibit a strong bias. Our limited knowledge of gene regulation mechanisms is immediately exposed when one attempts to explain this observation using current dogma. Our limited knowledge is further highlighted by an unrelated yet equally intriguing observation: disruption of IL-13 coding region was found to compromise IL-4 gene transcription, but only from the allele linked in cis to the mutant IL-13 allele (59).

Nucleosome Remodeling at the IL-12 p40 Promoter Most studies of chromatin at the IL-4 locus have focused on long-range events that can be monitored using DNase I hypersensitivity and DNA methylation assays. This initial focus is highly appropriate, given the primary objectives of understanding lineage commitment and the heritable regulation of transcription. Nevertheless, to elucidate the roles of specific DNA elements and DNA-binding protein in the

8 Feb 2002

10:32

AR

AR152-15.tex

AR152-15.SGM

LaTeX2e(2001/05/10)

P1: GSR

Annu. Rev. Immunol. 2002.20:427-462. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

CHROMATIN IN THE IMMUNE SYSTEM

445

remodeling of defined nucleosomes, higher resolution approaches are needed. To illustrate the mechanistic insight that can be gained from high-resolution studies, we briefly discuss studies of two other genes, the IL-12 p40 and IFN-β genes. Little is known about the contributions of distant control regions and global chromatin structure to the activation of the IL-12 p40 and IFN-β genes. In addition, the events involved in the transition to a decondensed chromatin structure during cell maturation have not been examined. Instead, the published studies have focused almost exclusively on events occurring at the promoter that contribute to the rapid induction of transcription during microbial infection of mature cells. The rationale for this starting point is that the promoter is most intimately involved in stimulating the formation of a transcription preinitiation complex. Furthermore, for most inducible genes, the promoter appears to be necessary and sufficient for transcriptional induction. Although distant control regions and global chromatin alterations are likely to be essential for highly efficient transcription and will undoubtedly play important roles in the formation of a preinitiation complex, the promoter comprises an essential, compact unit that is amenable to high-resolution analyses. The IL-12 p40 gene is activated in macrophages and other antigen presenting cells in response to microbial products, including LPS (60). The p40 protein is a subunit of the heterodimeric cytokine, IL-12, which serves as a bridge between innate and adaptive immune responses by stimulating the development of Th1 cells (60, 61). In transient transfection assays, a 400-bp fragment of the murine p40 promoter fused to a reporter gene is sufficient for transcriptional induction in macrophages through Toll-like receptor (TLR) signaling pathways (62–64). Extensive promoter mutant studies using the transient assay have identified five DNA elements that, in addition to the TATA box, are critical for promoter activity. Three of these elements bind members of the Rel, C/EBP, and AP-1 families (62, 63, 65). To analyze in detail the role of chromatin in the regulation of p40 promoter activity, the locations of nucleosomes in the vicinity of the promoter were first established (66). Nuclei from unactivated and activated cells were treated with micrococcal nuclease, which selectively introduces double-strand breaks in nucleosomal linker regions. The locations of the nuclease cleavage sites were then determined at low resolution by restriction enzyme cleavage followed by Southern blot analysis and at high resolution by ligation-mediated PCR (LM-PCR). The results revealed the existence of a positioned nucleosome extending from approximately −30 to −175 relative to the transcription start site. This nucleosome, referred to as nuclesome 1, encompasses the DNA elements identified in the transient transfection assay (Figure 3). An extended linker region (i.e., nucleosome-free region) was observed between −175 and −350, followed by an array of three more positioned nucleosomes. A DNase I and micrococcal nuclease hypersensitive site located upstream of these nucleosomes was observed only in macrophages (66). The presence of this strong hypersensitive site in mature macrophages prior to LPS stimulation provides evidence that the locus becomes decondensed during macrophage development.

1 Feb 2002

10:23

Annu. Rev. Immunol. 2002.20:427-462. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

446

AR

SMALE

AR152-15.tex

¥

AR152-15.SGM

LaTeX2e(2001/05/10)

P1: GSR

FISHER

Nucleosome remodeling upon macrophage stimulation with LPS was examined using a restriction enzyme accessibility assay. This assay monitors the efficiency of cleavage by a restriction enzyme at defined genomic sites upon incubation with nuclei from unstimulated and stimulated cells. Restriction sites within nucleosome 1 were cleaved much more efficiently in nuclei from stimulated cells than from unstimulated cells, providing evidence that nucleosome 1 is remodeled during transcriptional activation (66). In contrast, cleavages within regions flanking nucleosome 1 were not influenced by LPS stimulation, suggesting that remodeling is restricted to only one key nucleosome. Maximum accessibility was reached with only 1 h of stimulation, demonstrating that remodeling is rapid. However, protein synthesis was required for remodeling, as restriction enzyme accessibility did not increase when the cells were stimulated with LPS in the presence of cycloheximide or anisomycin (66). Interestingly ChIP results suggested that histone acetylation in the vicinity of the promoter increases only slightly during activation (64). This finding suggests that the locus may be fully acetylated in unstimulated cells and that an ATP-dependent remodeling complex may be responsible for the remodeling event detected using the restriction enzyme assay. The results described above suggest that activation of the IL-12 p40 gene requires an inducible and highly selective nucleosome remodeling event in addition to the induction of NF-κB, C/EBP, and AP-1 proteins. One possibility is that one or more of these transcription factors can rapidly access their binding sites within nucleosome 1 upon macrophage stimulation. These transcription factors could then recruit a nucleosome remodeling complex, which would provide access to the other DNA-binding proteins required for transcription. With this scenario, an understanding of p40 induction would rely primarily on an understanding of the induction of NF-κB, C/EBP, and AP-1 proteins; other inducible proteins would not be required for remodeling and promoter activity. An alternative model is that inducible remodeling may be independent of these transcription factors. Instead, remodeling may be required before any of these factors can gain access to their sites. With this model, other DNA-binding proteins would presumably be responsible for recruiting remodeling complex(es) to nucleosome 1. The binding sites for these proteins may have been missed in the original promoter mutant studies because they may have been unnecessary for gene activation in the transient transfection assay (41). To begin to distinguish between these two models, two initial strategies were employed. One strategy relied on the observation that IL-12 p40 transcription is greatly diminished in macrophages from c-Rel−/− mice (67). This result suggests that a c-Rel complex is likely to contribute to p40 activation by binding to the NF-κB site within the promoter. Interestingly, in c-Rel−/− macrophages that do not support transcription of the p40 locus, remodeling occurred at nearly wild-type levels, demonstrating that c-Rel is not necessary for remodeling (64). A second line of experiments supported the notion that the transcription factors identified using the transient transfection assay are insufficient and perhaps unnecessary for remodeling of nucleosome 1. Briefly, an analysis of macrophage cell

1 Feb 2002

10:23

AR

AR152-15.tex

AR152-15.SGM

LaTeX2e(2001/05/10)

P1: GSR

Annu. Rev. Immunol. 2002.20:427-462. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

CHROMATIN IN THE IMMUNE SYSTEM

447

lines containing a stably integrated IL-12 p40 promoter fused to a GFP reporter gene revealed that the proteins known to be required for promoter activity (i.e. NF– κB, C/EBP, and AP-1 proteins) are induced in all cells within an LPS-stimulated population (64). In contrast, the endogenous gene was induced in only a small fraction of the cells. The endogenous promoter was also cleaved by restriction enzymes in only a small fraction of cells, whereas the stably integrated promoters were cleaved efficiently both before and after stimulation. These results suggest that the NF-κB, C/EBP, and AP-1 proteins are efficiently induced and can readily access the stably integrated promoters, which contain constitutively accessible nucleosomes. However, these factors apparently are insufficient for remodeling of nucleosome 1 at the endogenous alleles. These results support the hypothesis that factors other than those required for transcription from transfected plasmids are required for the recruitment of remodeling complexes (Figure 3). Unfortunately, these experiments were indirect and are subject to a number of important caveats (64). Thus, although c-Rel is clearly unnecessary for remodeling, the involvement of the other factors remains uncertain. If remodeling of nucleosome 1 must, as hypothesized, precede the binding of NF-κB, C/EBP, and AP-1 proteins, identification of the proteins that recruit the remodeling complex will be critical for an understanding of p40 induction. Because there are no significant clues regarding the identities of these proteins, the DNA sequences within the locus that are required for inducible remodeling will first need to be identified. To achieve this goal, a transfection assay must be established that allows the promoter to assemble into an appropriate nucleosomal context, with nucleosome 1 properly positioned and inaccessible to restriction enzyme cleavage in unstimulated cells. To date, these properties have not been observed upon stable integration of moderately sized promoter fragments, suggesting that large genomic fragments or an episomal stable transfection assay will be required. If efficient and selective nucleosome 1 remodeling can be recapitulated with a transfected construct, a systematic mutant analysis can be employed to identify the DNA sequence elements required for inducible remodeling. After the DNA elements have been identified, the proteins that bind these elements can be characterized, eventually leading to the multiprotein complexes that are responsible for remodeling.

Nucleosome Remodeling at the IFN-β Promoter The IL-12 p40 analysis revealed that the five DNA elements important for promoter activity in a transient transfection assay are assembled into a positioned nucleosome at the endogenous locus. An analysis of c-Rel−/− macrophages and stably integrated IL-12 p40 promoter-reporter plasmids suggested that the DNA elements and inducible factors responsible for recruiting remodeling complexes to this nucleosome may have been missed in the transient transfection assay and may be difficult to identify. A discussion of the IFN-β promoter provides a contrasting view and an example of a different strategy through which nucleosome remodeling can contribute to transcriptional activation.

8 Feb 2002

10:33

Annu. Rev. Immunol. 2002.20:427-462. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

448

AR

SMALE

AR152-15.tex

¥

AR152-15.SGM

LaTeX2e(2001/05/10)

P1: GSR

FISHER

The IFN-β gene is induced in most cell types upon viral infection. Inducible transcription depends on a 55-bp region of the promoter located upstream of the TATA box (68, 69). In transient transfection experiments, virtually every nucleotide within this region is required for inducible promoter activity. The transcription factors that bind this region upon infection of HeLa cells with Sendai virus have been well characterized and include NF-κB, interferon regulatory factors (IRFs), an ATF-2/c-Jun heterodimer, and HMGI(Y). Through cooperative binding and major contributions from the architectural activities of HMGI(Y), these factors assemble into a highly stable complex, named an enhanceosome (68, 69). The surface of the intact enhanceosome interacts with the CREB-binding protein (CBP), a large, 265-kDa transcriptional coactivator that possesses histone acetyltransferase activity (68, 70, 71). Most of our knowledge of the role of chromatin in IFN-β induction has emerged from an elegant study that combined in vitro and in vivo approaches (72), although many earlier contributions were critical for the development of current models (68, 69). By formaldehyde cross-linking of nucleosomes to DNA, followed by micrococcal nuclease digestion and mapping of the nuclease-resistant DNA sequences, the critical 55-bp control region was found to be devoid of nucleosomes at the endogenous locus (72). Instead, positioned nucleosomes were identified immediately downstream (−15 to +132) and upstream (−268 to −118) of this region (Figure 4). Upon infection with Sendai virus, both of these nucleosomes were remodeled, as determined using a restriction enzyme accessibility assay. Increased restriction enzyme access was dependent on the proteins assembled into the enhanceosome. After obtaining this basic knowledge, Agalioti et al. (72) took advantage of the ChIP assay and one key feature of IFN-β gene induction to establish an apparent order of events that leads to inducible transcription (Figure 4). The key feature is that IFN-β gene transcription does not begin until 6 h after virus infection. This slow time-course allowed several important events to be placed into a temporal order. Using the ChIP assay, NF-κB and most of the other components of the enhanceosome bound the endogenous promoter by 2 h post-infection. Four hours after infection, CBP and RNA polymerase II associated with the promoter. The binding of these two proteins at a similar time, along with several other results, suggests that they exist as a complex and that the well-documented interaction between CBP and the enhanceosome is responsible for polymerase recruitment. Approximately 5 h post-infection, the Gcn5 histone acetyltransferase complex associated with the promoter. This association corresponded temporally with increased acetylation of histone H4, suggesting that Gcn5 is responsible for acetylation of this histone within the flanking nucleosomes. At 6 h post-infection, the BRG-1 subunit of the SWI/SNF ATP-dependent nucleosome remodeling complex associated with the promoter (72). This association corresponded temporally with increased access to restriction enzyme cleavage within the downstream nucleosome, suggesting that the restriction enzyme accessibility assay monitors the activity of an ATP-dependent remodeling complex rather

1 Feb 2002

10:23

AR

AR152-15.tex

AR152-15.SGM

LaTeX2e(2001/05/10)

P1: GSR

Annu. Rev. Immunol. 2002.20:427-462. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

CHROMATIN IN THE IMMUNE SYSTEM

449

than histone acetyltransferase activities. The final event in assembling the preinitiation complex at the IFN-β promoter appears to be the recruitment of transcription factor IID (TFIID), which contains the TATA-binding protein (TBP). This association also occurred at 6 h post-infection. The accumulated results suggest that these last few events occur in rapid succession. After the nucleosome encompassing the TATA box and transcription start site is remodeled by the SWI/SNF complex, TFIID gains access to this region and rapidly associates, followed by positioning of the previously recruited RNA polymerase II holoenzyme and transcription initiation. It is important to stress the value of the slow time-course of transcriptional activation for the success of this analysis. For genes that are induced more rapidly, including the IL-4 gene, IL-12 p40 gene, and most other primary and secondary response genes, this approach is unlikely to be feasible. Instead, it will be necessary to rely on genetic approaches, which, along with time-course studies, helped establish the order of events during activation of the yeast HO gene (73). For the IL-4 locus, the analysis of DNase I hypersensitivity in c-Maf−/− T cells provided an initial step toward establishing the order of events, by demonstrating that c-Maf acts downstream of the events involved in acquisition of the hypersensitive sites (37). For the IL-12 p40 promoter, the analysis of remodeling in c-Rel−/− macrophages provided the initial step, by demonstrating that c-Rel is primarily important for an event that follows nucleosome 1 remodeling (64). One weakness of using the ChIP assay to study the events involved in IFN-β activation is that it cannot establish a function for each factor and cannot establish the mechanism responsible for recruiting each factor. For this reason, Agalioti et al. (72) complemented the ChIP results with an elegant series of in vitro experiments employing DNA templates assembled into nucleosomes. Incubation of a promoter fragment with histones and nucleosome assembly factors resulted in the proper positioning of the nucleosome at the downstream position, presumably owing to the presence of positioning sequences within this DNA region. In vitro studies using these nucleosomal templates strengthened the ChIP results and provided important mechanistic insights (72). One key finding was that acetylation of histones within the downstream nucleosome stimulated recruitment of the BRG-1-containing SWI/SNF remodeling complex. SWI/SNF recruitment was accompanied by increased restriction enzyme access. When histone acetylation was blocked by a variety of different mechanisms, the SWI/SNF complex was not recruited and restriction enzyme access was not induced. These results strongly suggest that histone acetylation is essential for SWI/SNF association, helping to explain why SWI/SNF associated with the promoter in the ChIP experiments after the histone acetyltransferase complexes. Our current understanding of the IFN-β promoter greatly exceeds our understanding of any other metazoan promoter. Nevertheless, a number of questions remain to be answered. For example, what contributions are made by the posttranslational modification of enhanceosome components, and why does transcriptional activation require six hours? Recent studies of HMGI(Y) have begun to address

1 Feb 2002

10:23

Annu. Rev. Immunol. 2002.20:427-462. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

450

AR

SMALE

AR152-15.tex

¥

AR152-15.SGM

LaTeX2e(2001/05/10)

P1: GSR

FISHER

the first question (74), but the second question may be more difficult to answer. The events described above are unlikely to require six hours to complete, as many other genes that are induced more rapidly are likely to require similar events prior to transcription initiation. The slow time-course of IFN-β induction may be important for an optimal antiviral response, but the biological reason and underlying mechanism are not immediately apparent. Another key question is the following: What is the significance of the different activation strategies employed by the IFN-β and IL-12 p40 promoter? Although nucleosome remodeling is involved in the induction of both promoters, the key transcription factor binding sites within the IFN-β promoter are nucleosome-free, whereas the key binding sites within the IL-12 p40 promoter are assembled into a positioned nucleosome. Another important and perhaps related difference is that an extremely stable enhanceosome forms at the IFN-β promoter, whereas cooperative binding by factors at the IL-12 p40 promoter has been difficult to detect. One possibility is that the two strategies represent different mechanisms for preventing the aberrant activation of transcription, which has the potential to occur when only a subset of the key transcription factors are induced. Aberrant activation of the IL-12 p40 may be suppressed by nucleosome 1, which may prevent NF-κB, C/EBP, and AP-1 proteins from binding and activating transcription in the absence of a tightly regulated nucleosome remodeling event. In contrast, aberrant activation of the IFN-β promoter may be prevented by the requirement for highly cooperative binding of the enhancesome components. The binding affinities of individual factors may be too low to result in transcriptional activation, ensuring that the gene will be transcribed only when all of the enhanceosome factors have been induced.

HERITABLE GENE INACTIVATION In recent years, there has been tremendous interest in defining the molecular events that underlie gene activation and the onset of transcription. By comparison, the molecular basis of gene silencing and heritable gene inactivation has received considerably less attention. This is surprising, not only because a large proportion of genes are inactive in the cells of higher eukaryotes, but also because heritable gene silencing is likely to be important for many developmental processes. For example, as lymphocytes differentiate from hematopoietic precursors, they undergo a progressive loss in their ability to switch to alternate cell types. This directional loss (or restriction) of lineage potential is probably more consistent with a progressive silencing of the genome than with its sequential activation. This point is elegantly underscored by the demonstration that the transcription factor Pax-5 is crucial for B cell commitment because it is required for the downregulation of genes defining alternative lineage options (75, 76). Although the relationship between chromatin and gene activation was discussed above by focusing on specific model genes, it is not possible at this stage to use

1 Feb 2002

10:23

AR

AR152-15.tex

AR152-15.SGM

LaTeX2e(2001/05/10)

P1: GSR

Annu. Rev. Immunol. 2002.20:427-462. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

CHROMATIN IN THE IMMUNE SYSTEM

451

a parallel approach to discuss gene silencing. A few control regions have been characterized that are necessary for the heritable silencing of genes in the immune system, the most prominent of which is the CD4 silencer (77, 78). However, no gene silencing event has been studied from the perspective of chromatin structure in a manner that could be considered analogous to the studies of IL-4, IL-12 p40, and IFN-β gene activation. For this reason, the following sections describe only briefly some of the key factors involved in heritable gene silencing in eukaryotes. In addition to summarizing current knowledge of gene silencing mechanisms, these sections provide further insight into mechanisms of heritable gene activation, as some of the topics relate to both processes. Although much of this knowledge originated from studies in yeast and flies, converging evidence suggests that homologous factors are critical for the normal development and functioning of cells of the immune system.

Polycomb-Group (Pc-G) and Trithorax-Group(Trx-G) Proteins Although cellular identity is “read out” primarily at the level of transcriptional responses, many of the key regulators that set up transcriptional competence are expressed only fleetingly at early stages of development. In order for this initial repertoire of gene activation and repression to be remembered by cells at later stages, a second set of factors (encoded by the Polycomb and Trithorax group genes) are necessary to stabilize and maintain these expression profiles. For example, in Drosophila melanogaster, products of the gap and pair rule genes initially specify the correct expression of the homeotic Bithorax and Antennapedia complexes (79). However, to maintain appropriate expression throughout development, the concerted action of both Pc-G and Trx-G gene products is required (80, 81). These products were largely identified in genetic screens and have provided something of an enigma for molecular biologists. They act in complexes, exert effects over large distances, are sensitive to dose, and yet with a few exceptions (GAGA, Zeste and Pho) do not directly interface with DNA. Evidence that the mammalian counterparts of these genes may have a similar function in lymphocytes has come through the targeted disruption or overexpression of several Pc-G members in mice. Mutant mice lacking Bmi-1 or Mel-18 (both orthologues of posterior sex combs), Rae28, or M33 (orthologues of polyhomeotic and polycomb, respectively) show retarded growth and homeotic transformations that parallel the phenotypes of mutant flies. In addition, the mutant mice exhibit profound abnormalities within the lymphoid compartment (82–85). In particular, the spleen and thymus of homozygous null mice are small and contain lymphocytes that are greatly impaired in their proliferative responses to mitogens. Conversely, overexpression of Bmi-1 predisposed mice to lymphoma (86), consistent with a role for Pc-G proteins in cell cycle control in lymphocytes. Although current studies do not allow us to define exactly how Pc-G proteins influence lymphocyte proliferation, demonstrations that loss of the INK4A tumor suppressor can partially complement the loss of Bmi-1 (87), and that Mel-18 is involved in c-myc

1 Feb 2002

10:23

Annu. Rev. Immunol. 2002.20:427-462. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

452

AR

SMALE

AR152-15.tex

¥

AR152-15.SGM

LaTeX2e(2001/05/10)

P1: GSR

FISHER

regulation in lymphocytes (88), suggest that further insights into these processes will be realized in the longer term. The central issue of how Pc-G proteins recognize transcriptionally silent states and subsequently propagate transcriptional inactivity through DNA replication and cell division is not understood. Whether Pc-G proteins recognize molecular differences in the composition of transcribed and nontranscribed genes, such as changes in the histone content (or modifications), nucleosome positioning, the binding of associated factors, or the extent of chromatin compaction, is simply not known. Similarly, whether Pc-G proteins repress gene transcription by directly interfering with the transcriptional machinery, recruiting histone deacetylases, or preventing the remodeling of a locus prior to gene activation, are questions for which we presently have only fragmentary answers.

Pericentromeric Heterochromatin Much of what we do know about facultative heterochromatin comes from experimental studies of constitutive heterochromatin and, in particular, the heterochromatin that surrounds functional centromeres. Pericentromeric DNA is characterized by long tracts of tandemly repeated, highly methylated satellite DNA. Although the DNA sequence and length of repeats are highly variable between organisms, several structural features appear consistent, including the binding of heterochromatin proteins (such as HP1) and an underacetylation of histone H4 (89). This environment is transcriptionally unfavorable, and the integration of euchromatic transgenes into pericentromeric domains often results in their silencing. Model systems in which a disabled LCR is introduced into heterochromatin have provided evidence of position effect variegation (PEV) in mammals. These findings are consistent with demonstrations of PEV in flies, in which a variable proportion of cells express or do not express in a heritable manner the integrated transgene. In the non-expressing population, the transgene adopts the chromatin characteristics of the neighboring region, while in expressing cells, the transgene can apparently override these repressive constraints. An appealing example of PEV is provided by mice carrying the human CD2 transgene (90), in which T cells expressing the transgene can be purified by FACS and their properties examined in detail. These and similar models for the human β-globin locus (91) have been used to characterize the contribution of individual heterochromatin components for gene silencing. For example, by intercrossing human CD2 transgenic mice with mice overexpressing mouse HP1, a higher proportion of cells that fail to express the transgene are generated (92). Although this result confirms that HP1 can contribute to functional silencing within heterochromatin, the studies further suggested that whether HP1 enhances or suppresses expression depends on the exact nature of the neighboring heterochromatic environment. The apparent discrepancy in the behavior of transgenes integrated within heterochromatin in pericentromeric and noncentromeric locations awaits a comprehensive molecular explanation.

1 Feb 2002

10:23

AR

AR152-15.tex

AR152-15.SGM

LaTeX2e(2001/05/10)

P1: GSR

Annu. Rev. Immunol. 2002.20:427-462. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

CHROMATIN IN THE IMMUNE SYSTEM

453

Evidence from studies in flies has also shown that unlinked genes that are moved close to heterochromatin (in trans) can be silenced. The mutant allele of the brown eye color locus of Drosophila (bwD) contains an extensive (>1 mb) block of satellite AAGAG DNA inserted into the coding region of the locus. In heterozygotes (bw+/bwD), silencing of the wild-type allele is promoted by the somatic pairing of the bw+ allele with its dominant allele and the recruitment of both alleles to constitutive heterochromatin bearing the same AAGAG repeat. In this example, the normal bw+ allele is silenced by virtue of its physical recruitment to pericentric heterochromatin (93). Modifiers of PEV that increase the proximity of this interaction also enhance the silencing of the bw+ allele (94), providing direct evidence that these components can influence gene recruitment within the nucleus.

Ikaros and Gene Silencing in Lymphocytes The observation that the DNA-binding protein Ikaros is associated with several transcriptionally inactive genes, and that these loci are recruited to centromeric DNA in cycling lymphocytes, has prompted a series of speculations about its role (95, 96). Ikaros was originally identified as a protein that binds the D’ element within the murine terminal transferase (TdT ) gene promoter and an element within the murine CD3δ enhancer (97, 98). Although we do not yet know how Ikaros regulates gene expression or even why it is so vital for the normal development of T, B, and NK cells (99), it is likely that Ikaros has a role beyond that of a conventional transcriptional activator. For example, Ikaros proteins compete with the activator protein Elf-1 (an Ets family member) for binding to the D’ region of the TdT promoter (Figure 5); mutations that disrupt Ikaros binding prevent downregulation of the TdT gene during T cell differentiation (100). Parallel studies examining the contribution of Ikaros binding motifs in the promoter for the λ5 gene have also shown the importance of Ikaros for gene silencing (101). In this study employing transgenic mice, mutation of an Ikaros binding site (Figure 5) resulted in the failure of the transgene to be silenced appropriately in mature B cells. Interestingly, integration of this mutant transgene directly into pericentric heterochromatin did not restore normal silencing. These data provide evidence that Ikaros proteins can initiate the gene silencing process. The establishment of heritable gene silencing may require several sequential steps, which may also require the functions of Ikaros proteins. It is conceivable that the recruitment of inactive genes close to centromeric heterochromatin serves to stabilize gene repression, either on the basis of energetic considerations or because this compartment contains components (such as histone deactylases and methyltransferases) that are involved in propagating the inactive state. In this respect, it is intriguing that Ikaros binds directly to major satellite DNA (102) and associates with the NuRD complex, which contains histone deacetylases and nucleosome remodeling factors (103). Since Ikaros multimers appear to interact with target genes, centromeric DNA, and chromatin modifiers, it is conceivable that

1 Feb 2002

10:23

Annu. Rev. Immunol. 2002.20:427-462. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

454

AR

SMALE

AR152-15.tex

¥

AR152-15.SGM

LaTeX2e(2001/05/10)

P1: GSR

FISHER

Figure 5 Working model for the downregulation and heritable inactivation of candidate Ikaros target genes in developing lymphocytes.

Ikaros mediates the conversion of candidate genes from transiently to permanently repressed states. Although such a role would be consistent with the observation that the rag-1 and TdT genes are recruited to Ikaros-containing domains in thymocytes during heritable, but not transient, silencing (96), no formal proof for this hypothesis has yet been shown.

Heritable Silencing and the Formation of Stable Epigenetic Imprints Although most of the details of how cellular memory operates are still unknown, several conceptual models have been proposed (36). These models have assumed that intrinsic, self-templating mechanisms underlie the propagation of chromatin states and that active and inactive chromatin structures are “tagged” by a variety of means. Accordingly, as described above, hypotheses have been proposed in which active and inactive chromatin regions are functionally segregated by modification of the DNA itself (by methylation) or of associated proteins such as histones (by methylation, acetylation, or other posttranslational modifications). DNA methylation can be propagated through DNA replication by DNMT1, which, as described earlier, selectively methylates hemimethylated DNA strands during S phase. Histone modifications or other protein tags could be propagated by differences in the timing or location of DNA synthesis. The idea that the replication timing and nuclear location of active and inactive chromatin might be distinct has been supported by correlative evidence. However, it remains unclear whether such a generalized model could reasonably accommodate the patchwork of active

1 Feb 2002

10:23

AR

AR152-15.tex

AR152-15.SGM

LaTeX2e(2001/05/10)

P1: GSR

Annu. Rev. Immunol. 2002.20:427-462. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

CHROMATIN IN THE IMMUNE SYSTEM

455

and inactive genes resident along the autosomal chromosomes of differentiated cells. Although progress in this area has been slow, complementary research in understanding the extent and diversity of histone modifications has led to the proposal that covalent modifications to these proteins could form the basis of a novel epigenetic code (4). A string of related reports has shown that the acetylation, phosporylation, and methylation of histone tails can profoundly affect the recruitment of chromatin-associated proteins and transcriptional components (Figure 6). For example the lysine acetylation of histone H4 induces an increased affinity for bromodomain proteins such as TAFII250 and P/CAF (104). Because these proteins have intrinsic istone acetyltransferase activities, this recruitment may serve to enhance transcriptional activity at a histone tagged site. Conversely, methylation of histone H3 at lysine 9 by the methyltransferase Suv39H1 targets binding of the heterochromatin component HP-1, a protein that is associated with transcriptional repression (Figure 6) (105). Interestingly, whereas methylation of histone H3 at lysine 9 is a characteristic of inactive heterochromatin (such as that surrounding mouse centromeres), methylation of lysine at position 4 is a selective feature of active domains (Figure 6). This observation underscores the repertoire and complexity of this emerging epigenetic code. Our future challenges will be to clarify the links between these molecular tags and upstream mediators such as Pc-G and

Figure 6 Summary of events likely to contribute to the establishment of stable epigenetic imprints.

1 Feb 2002

10:23

456

AR

SMALE

AR152-15.tex

¥

AR152-15.SGM

LaTeX2e(2001/05/10)

P1: GSR

FISHER

Trx-G, and also to examine in greater detail their relevance to the coordinate regulation of complex loci.

Annu. Rev. Immunol. 2002.20:427-462. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

SUMMARY AND FUTURE GOALS We have described examples in which chromatin structure appears to make important contributions to gene regulation in the immune system. Instead of listing the many instances in which transcription factors have been reported to interact with chromatin remodeling and histone modification complexes, we have focused on specific genes that are being subjected to systematic chromatin analyses. Although a considerable amount of effort was required to develop these types of model systems, similar approaches will need to be applied to other genes to uncover the mechanisms by which they are regulated by key transcription factors. Although a long-term goal will be to elucidate the functions of specific trans-acting factors in modulating chromatin structure, an analysis of cis-acting sequences will be of equal importance for the next series of advances. For example, to determine how the IL-4 locus is decondensed during Th2 development, it will be necessary to identify the control regions and specific DNA elements required for this process. An exclusive focus on the characterization of trans-acting factors would be insufficient because of the inherent difficulty distinguishing direct from indirect effects upon overexpression or underexpression of a factor. Similarly, at the IL-12 p40 locus, the specific DNA elements required for nucleosome remodeling will need to be identified before the critical trans-acting factors can be pursued. The future goals of the IFN-β chromatin analysis appear to be quite different because all of the components of the stable enhanceosome appear to act in concert to stimulate remodeling of the downstream nucleosome. It is possible that an analogous situation will be found at the IL-4 locus, such that every DNA element within one or more control regions will be necessary for chromatin decondensation. With this scenario, the functions of specific transcription factors would be much less significant than the concerted functions of a specific combination of factors. In addition to uncovering mechanisms responsible for chromatin decondensation and nucleosome remodeling at specific loci, a major goal will be to elucidate the mechanisms that contribute to the heritable propagation of transcriptionally competent and silent states. As described above, recent advances in dissecting the histone modification code and the regulation of DNA methylation have generated considerable optimism that this goal will be realized in the foreseeable future. The current studies of gene regulation in the immune system are likely to make important contributions to this effort. The Th1/Th2 differentiation system is particularly well-suited for studies of the events involved in heritable lineage decisions: The culture conditions and molecular events required for the in vitro differentiation of T helper cells have been relatively well established, the differentiation process is amenable to analysis using gene disruption technologies, and it has received, quite appropriately, the attention of a large number of laboratories that are

1 Feb 2002

10:23

AR

AR152-15.tex

AR152-15.SGM

LaTeX2e(2001/05/10)

P1: GSR

CHROMATIN IN THE IMMUNE SYSTEM

457

consistently performing elegant and ambitious experiments. The groundbreaking studies of allelic bias nicely exemplify the power of this system and the potential for contributions that greatly advance our broader understanding of gene regulation mechanisms in mammalian cells. ACKNOWLEDGMENTS

Annu. Rev. Immunol. 2002.20:427-462. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

We thank Jane Grogan, Richard Locksley, Matthias Merkenschlager, and Anjana Rao for valuable comments and critical reading of the text. Visit the Annual Reviews home page at www.annualreviews.org

LITERATURE CITED 1. Van Holde KE. 1989. Chromatin. New York: Springer-Verlag 2. Kingston RE, Narlikar GJ. 1999. ATPdependent remodeling and acetylation as regulators of chromatin fluidity. Genes Dev. 13:2339–42 3. Peterson CL, Workman JL. 2000. Promoter targeting and chromatin remodeling by the SWI/SNF complex. Curr. Opin. Genet. Dev. 10:187–92 4. Jenuwein T, Allis CD. 2001. Translating the histone code. Science 293:1074–80 5. Wallrath LL. 1998. Unfolding the mysteries of heterochromatin. Curr. Opin. Genet. Dev. 8:147–53 6. Cockell M, Gasser SM. 1999. Nuclear compartments and gene regulation. Curr. Opin. Genet. Dev. 9:199–205 7. Belmont AS, Dietzel S, Nye AC, Strukuv YG, Tumbar T. 1999. Large scale chromatin structure and function. Curr. Opin. Cell Biol. 11:307–11 8. Jones PA, Takai D. 2001. The role of DNA methylation in mammalian epigenetics. Science 293:1068–70 9. Sleckman BP, Bassing CH, Bardon CG, Okada A, Khor B, Bories JC, Monroe R, Alt FW. 1998. Accessibility control of variable region gene assembly during T-cell development. Immunol. Rev. 165: 121–30 10. Hernandez-Munain C, McMurry MT, Krangel MS. 1999. Regulation of chro-

11.

12.

13.

14.

15.

16.

17.

matin accessibility for V(D)J recombination. Cold Spr. Harb. Symp. Quant. Biol. 64:183–89 Roth DB, Roth SY. 2000. Unequal access: regulating V(D)J recombination through chromatin remodeling. Cell 103:699–702 Glimcher LH, Murphy KM. 2000. Lineage commitment in the immune system: the T helper lymphocyte grows up. Genes Dev. 14:1693–711 O’Garra A, Arai N. 2000. The molecular basis of T helper 1 and T helper 2 cell differentiation. Trends Cell Biol. 10:542–50 Grogan JL, Mohrs M, Harmon B, Lacy DA, Sedat JW, Locksley RM. 2001. Early transcription and silencing of cytokine genes underlie polarization of T helper cell subsets. Immunity 14:205–15 Pannetier C, Hu-Li J, Paul WE. 1999. Bias in the expression of IL-4 alleles: the use of T cells from a GFP knockin mouse. Cold Spr. Harb. Symp. Quant. Biol. 64:599–602 Bird JJ, Brown DR, Mullen AC, Moskowitz NH, Mahowald MA, Sider JR, Gajewski TF, Wang CR, Reiner SL. 1998. Helper T cell differentiation is controlled by the cell cycle. Immunity 9:229–37 Ben-Sasson SZ, Gerstel R, Hu-Li J, Paul WE. 2001. Cell division is not a “clock” measuring acquisition of competence to produce IFN-gamma or IL-4. J. Immunol. 166:112–20

1 Feb 2002

10:23

Annu. Rev. Immunol. 2002.20:427-462. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

458

AR

SMALE

AR152-15.tex

¥

AR152-15.SGM

LaTeX2e(2001/05/10)

P1: GSR

FISHER

18. Loots GG, Locksley RM, Blankespoor CM, Wang ZE, Miller W, Rubin EM, Frazer KA. 2000. Identification of a coordinate regulator of interleukins 4, 13, and 5 by cross-species sequence comparisons. Science 288:136–40 19. Bix M, Locksley RM. 1998. Independent and epigenetic regulation of the interleukin-4 alleles in CD4+ T cells. Science 281:1352–54 20. Rivi`ere I, Sunshine MJ, Littman DR. 1998. Regulation of IL-4 expression by activation of individual alleles. Immunity 9:217–28 21. Hu-Li J, Pannetier C, Guo L, L¨ohning M, Gu H, Watson C, Assenmacher M, Radbruch A, Paul WE. 2001. Regulation of expression of IL-4 alleles: analysis using a chimeric GFP/IL-4 gene. Immunity 14:1–11 22. Kelly BL, Locksley RM. 2000. Coordinate regulation of the IL-4, IL-13, and IL5 cytokine cluster in Th2 clones revealed by allelic expression patterns. J. Immunol. 165:2982–86 23. Sekinger EA, Gross DS. 2001. Silenced chromatin is permissive to activator binding and PIC recruitment. Cell 105:403–14 24. Lacy DA, Wang ZE, Symula DJ, McArthur CJ, Rubin EM, Frazer KA, Locksley RM. 2000. Faithful expression of the human 5q31 cytokine cluster in transgenic mice. J. Immunol. 164:4569–74 25. Mohrs M, Blankespoor CM, Wang Z-E, Loots GG, Afzal V, Hadeiba H, Shinkai K, Rubin EM, Locksley RM. 2001. Deletion of a coordinate regulator of type 2 cytokine expression in mice. Nat. Immunol. 2:842–47 26. Weintraub H. 1988. Formation of stable transcription complexes as assayed by analysis of individual templates. Proc. Natl. Acad. Sci. USA 85:5819–23 27. Ellwood K, Chi T, Huang W, Mitsouras K, Carey M. 1998. Cooperative assembly of RNA polymerase II transcription complexes. Cold Spr. Harb. Symp. Quant. Biol. 63:253–61

28. Weintraub H, Groudine M. 1976. Chromosomal subunits in active genes have an altered conformation. Science 193:848– 56 29. Hebbes TR, Clayton AL, Thorne AW, Crane-Robinson C. 1994. Core histone hyperacetylation co-maps with generalized DNase I sensitivity in the chicken β-globin chromosomal domain. EMBO J. 13:1823–30 30. Schubeler D, Francastel C, Cimbora DM, Reik A, Martin DI, Groudine M. 2000. Nuclear localization and histone acetylation: a pathway for chromatin opening and transcriptional activation of the human beta-globin locus. Genes Dev. 14:940–50 31. Carey M, Smale ST. 2000. Transcriptional regulation in eukaryotes: concepts, strategies, and techniques. Cold Spring Harbor, NY: Cold Spr. Harb. Lab. Press 32. Wu C, Bingham PM, Livak KJ, Holmgren R, Elgin SC. 1979. The chromatin structure of specific genes: I. Evidence for higher order domains of defined DNA sequence. Cell 16:807–14 33. Takemoto N, Koyano-Nakagawa N, Yokota T, Arai N, Miyatake S, Arai K. 1998. Th2-specific DNase I-hypersensitive sites in the murine IL-13 and IL4 intergenic region. Int. Immunol. 10: 1981–85 34. Agarwal S, Rao A. 1998. Modulation of chromatin structure regulates cytokine gene expression during T cell differentiation. Immunity 9:765–75 35. Kubo M, Ransom J, Webb D, Hashimoto Y, Tada T, Nakayama T. 1997. T-cell subset-specific expression of the IL-4 gene is regulated by a silencer element and STAT6. EMBO J. 16:4007–20 36. Riggs AD, Pfeifer GP. 1992. X-chromosome inactivation and cell memory. Trends Genet. 8:169–74 37. Agarwal S, Avni O, Rao A. 2000. Celltype-restricted binding of the transcription factor NFAT to a distal IL-4 enhancer in vivo. Immunity 12:643–52 38. Todd MD, Grusby MJ, Lederer JA, Lacy

1 Feb 2002

10:23

AR

AR152-15.tex

AR152-15.SGM

LaTeX2e(2001/05/10)

P1: GSR

CHROMATIN IN THE IMMUNE SYSTEM

39.

Annu. Rev. Immunol. 2002.20:427-462. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

40.

41.

42.

43.

44.

45.

46.

47.

48.

49.

E, Lichtman AH, Glimcher LH. 1993. Transcription of the interleukin 4 gene is regulated by multiple promoter elements. J. Exp. Med. 177:1663–74 Wenner CA, Szabo SJ, Murphy KM. 1997. Identification of IL-4 promoter elements conferring Th2-restricted expression during T helper cell subset development. J. Immunol. 158:765–73 Lee GR, Fields PE, Flavell RA. 2001. Regulation of IL-4 gene expression by distal regulatory elements and GATA-3 at the chromatin level. Immunity 14:447–59 Smith CL, Hager GL. 1997. Transcriptional regulation of mammalian genes in vivo: a tale of two templates. J. Biol. Chem. 272:27493–96 Bell AC, West AG, Felsenfeld G. 2001. Insulators and boundaries: versatile regulatory elements in the eukaryotic genome. Science 291:447–50 Grosveld F. 1999. Activation by locus control regions? Curr. Opin. Genet. Dev. 9:152–57 Kellum R, Schedl P. 1991. A positioneffect assay for boundaries of higher order chromosomal domains. Cell 64:941–50 Ernst P, Smale ST. 1995. Combinatorial regulation of transcription I: General aspects of transcriptional control. Immunity 2:311–19 Zheng W-P, Flavell RA. 1997. The transcription factor GATA-3 is necessary and sufficient for Th2 cytokine gene expression in CD4 T cells. Cell 89:587–96 Ho I-C, Hodge MR, Rooney JW, Glimcher LH. 1996. The proto-oncogene c-maf is responsible for tissue-specific expression of interleukin-4. Cell 85:973–83 Hodge MR, Ranger AM, de la Brousse F, Hoey T, Grusby T, Glimcher LH. 1996. NFAT-driven interleukin-4 transcription potentiated by NIP45. Science 274:1903– 5 Weiss DL, Brown MA. 2001. Regulation of IL-4 production in mast cells: a paradigm for cell-type-specific gene expression. Immunol. Rev. 179:35–47

459

50. Ouyang W, L¨ohning M, Gao Z, Assenmacher M, Ranganath S, Radbruch A, Murphy KM. 2000. Stat6-independent GATA-3 autoactivation directs IL-4independent Th2 development and commitment. Immunity 12:27–37 51. Lee HJ, Takemoto N, Kurata H, Kamogawa Y, Miyatake S, O’Garra A, Arai N. 2000. GATA-3 induces T helper cell type 2 (Th2) cytokine expression and chromatin remodeling in committed Th1 cells. J. Exp. Med. 192:105–15 52. Takemoto N, Kamogawa Y, Jun Lee H, Kurata H, Arai KI, O’Garra A, Arai N, Miyatake S. 2000. Cutting edge: chromatin remodeling at the IL-4/IL-13 intergenic regulatory region for Th2specific cytokine gene cluster. J. Immunol. 165:6687–91 53. Szabo SJ, Kim ST, Costa GL, Zhang X, Fathman GC, Glimcher LH. 2000. A novel transcription factor, T-bet, directs Th1 lineage commitment. Cell 100:655– 69 54. Mullen AC, High FA, Hutchins AS, Lee HW, Villarino AV, Livingston DM, Kung AL, Cereb N, Yao TP, Yang SY, Reiner SL. 2001. Role of T-bet in commitment of TH1 cells before IL-12-dependent selection. Science 292:1907–10 55. Blobel GA, Nakajima T, Eckner R, Montminy M, Orkin SH. 1998. CREBbinding protein cooperates with transcription factor GATA-1 and is required for erythroid differentiation. Proc. Natl. Acad. Sci. USA 95:2061–66 56. Kim J, Ho IC, Grusby M, Glimcher LH. 1999. The transcription factor c-Maf controls the production of IL-4 but not other Th2 cytokines. Immunity 10:745–51 57. Holl¨ander GA, Zuklys S, Morel C, Mizoguchi E, Mobisson K, Simpson S, Terhorst C, Wishart W, Golan DE, Bhan AK, Burakoff SJ. 1998. Monoallelic expression of the interleukin-2 locus. Science 279:2118–21 58. Held W, Roland J, Raulet DH. 1995. Allelic exclusion of Ly49-family genes

1 Feb 2002

10:23

460

59.

Annu. Rev. Immunol. 2002.20:427-462. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

60.

61.

62.

63.

64.

65.

66.

67.

AR

SMALE

AR152-15.tex

¥

AR152-15.SGM

LaTeX2e(2001/05/10)

P1: GSR

FISHER

encoding class I MHC-specific receptors on NK cells. Nature 376:355–58 Guo L, Hu-Li J, Zhu J, Pannetier C, Watson C, McKenzie GJ, McKenzie AN, Paul WE. 2001. Disrupting Il13 impairs production of IL-4 specified by the linked allele. Nat. Immunol. 2:461–66 Trinchieri G. 1995. Interleukin-12: a proinflammatory cytokine with immunoregulatory functions that bridge innate resistance and antigen-specific adaptive immunity. Annu. Rev. Immunol. 13:251– 76 Abbas AK, Murphy KM, Sher A. 1996. Functional diversity of helper T lymphocytes. Nature 383:787–93 Murphy TL, Cleveland MG, Kulesza P, Magram J, Murphy KM. 1995. Regulation of interleukin 12 p40 expression through an NF-B half-site. Mol. Cell. Biol. 15:5258–67 Plevy SE, Gemberling JHM, Hsu S, Dorner AJ, Smale ST. 1997. Multiple control elements mediate activation of the murine and human interleukin 12 p40 promoters: evidence of functional synergy between C/EBP and Rel proteins. Mol. Cell. Biol. 17:4572–88 Weinmann AS, Mitchell DM, Sanjabi S, Bradley MN, Hoffmann A, Liou H-C, Smale ST. 2001. Nucleosome remodeling at the IL-12 p40 promoter is a TLRdependent, Rel-independent event. Nature Immunol. 2:51–57 Zhu C, Gagnidze K, Gemberling JH, Plevy SE. 2001. Characterization of an activation protein-1-binding site in the murine interleukin-12 p40 promoter. Demonstration of novel functional elements by a reductionist approach. J. Biol. Chem. 276:18,519–28 Weinmann AS, Plevy SE, Smale ST. 1999. Rapid and selective remodeling of a positioned nucleosome during the induction of IL-12 p40 transcription. Immunity 11:665–75 Sanjabi S, Hoffmann A, Liou H-C, Baltimore D, Smale ST. 2000. Selective

68.

69.

70.

71.

72.

73.

74.

75.

76.

requirement for c-Rel during IL-12 p40 gene induction in macrophages. Proc. Natl. Acad. Sci. USA 97:12705–10 Maniatis T, Falvo JV, Kim TH, Kim TK, Lin CH, Parekh BS, Wathelet MG. 1998. Structure and function of the interferonβ enhanceosome. Cold Spr. Harb. Symp. Quant. Biol. 63:609–20 Munshi N, Yie J, Merika M, Senger K, Lomvardas S, Agalioti T, Thanos D. 1999. The IFN-β enhancer: a paradigm for understanding activation and repression of inducible gene expression. Cold Spr. Harb. Symp. Quant. Biol. 64:149– 59 Kwok RP, Lundblad JR, Chrivia JC, Richards JP, Bachinger HP, Brennan RG, Roberts SGE, Green MR, Goodman RH. 1994. Nuclear protein CBP is a coactivator for the transcription factor CREB. Nature 370:223–26 Ogryzko VV, Schiltz RL, Russanova V, Howard BH, Nakatani Y. 1996. The transcriptional coactivators p300 and CBP are histone acetyltransferases. Cell 87:953– 59 Agalioti T, Lomvardas S, Parekh B, Yie J, Maniatis T, Thanos D. 2000. Ordered recruitment of chromatin modifying and general transcription factors to the IFN-β promoter. Cell 103:667–78 Cosma MP, Tanaka T, Nasmyth K. 1999. Ordered recruitment of transcription and chromatin remodeling factors to a cell cycle and developmentally regulated promoter. Cell 97:299–311 Munshi N, Agalioti T, Lomvardas S, Merika M, Chen G, Thanos D. 2001. Coordination of a transcriptional switch by HMGI(Y) acetylation. Science 293:1133– 36 Nutt SL, Heavey B, Rolink AG, Busslinger M. 1999. Commitment to the BLymphoid lineage depends on the transcription factor Pax5. Nature 401:538–39 Busslinger M, Nutt SL, Rolink AG. 2000. Lineage commitment in lymphopoiesis. Curr. Opin. Immunol. 12:151–58

1 Feb 2002

10:23

AR

AR152-15.tex

AR152-15.SGM

LaTeX2e(2001/05/10)

P1: GSR

Annu. Rev. Immunol. 2002.20:427-462. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

CHROMATIN IN THE IMMUNE SYSTEM 77. Siu G, Wurster AL, Duncan DD, Soliman TM, Hedrick SM. 1994. A transcriptional silencer controls the developmental expression of the CD4 gene. EMBO J. 13:3570–79 78. Sawada S, Scarborough JD, Killeen N, Littman DR. 1994. A lineage-specific transcriptional silencer regulates CD4 gene expression during T lymphocyte development. Cell 77:917–29 79. Akam M. 1987. The molecular basis for metameric pattern in the Drosophila embryo. Development 101:1–22 80. Kennison JA. 1995. The polycomb and trithorax group proteins of drosophila; trans-regulators of homeotic gene function. Annu. Rev. Genet. 29:289–303 81. Francis NJ, Kingston RE. 2001. Mechanisms of transcriptional memory. Mol. Cell. Biol. 2:409–21 82. Van der Lugt NM, Domen J, Linders K, van Roon M, Robanus-Maandag E, teRiele H, van der Valk M, Deschamps J, Sofroniew M, van Lohuizen M, Berns A. 1994. Posterior transformation, neurological abnormalities and sever hematopoietic defects in mice with a targeted deletion of the bmi-1 proto-oncogene. Genes Dev. 8:757–69 83. Akasaka T, Kanno M, Balling R, Mieza MA, Taniguchi M, Koseki H. 1996. A role for mel-18, a Polycomb group-related vertebrate gene, during the anterposterior specification of the axial skeleton. Development 122:1513–22 84. Core N, Bel S, Gaunt SJ, AurrandLions M, Pearce J, Fisher A, Djabali M. 1997. Altered cellular proliferation and mesoderm patterning in Polycomb-M33deficient mice. Development 124:721– 29 85. Takihara Y, Tomotsune D, Shirai M, Katoh-fukui Y, Nishii K, Motaleb MA, Nomura M, Tsuchiya R, Fujita Y, Shibata Y, Higashinakagawa T, Shimada K. 1997. Targeted disruption of the mouse homologue of the Drosophila polyhomeotic gene leads to altered anteroposterior pat-

86.

87.

88.

89.

90.

91.

92.

93.

461

terning and neural crest defects. Development 124:3673–82 Alkema MJ, van der Lugt NMT, Bobeldijk RC, Berns A, van Lohuizen M. 1995. Transformation of axial skeleton due to overexpression of bmi-1 in transgenic mice. Nature 374:724–27 Jacobs JJ, Kieboom K, Marino S, DePinho RA, van Lohuizen M. 1999. The oncogene and Polycomb-group gene bmi1 regulates cell proliferation and senescence through the ink4a locus. Nature 397:164–68 Tetsu O, Ishihara H, Kanno R, Kamiyasu M, Innoue H, Tokuhisa T, Taniguchi M, Kanno M. 1998. Mel-18 negatively regulates cell cycle progression upon B cell antigen receptor stimulation through a cascade leading to c-myc/cdc25. Immunity 9:439–48 Choo KHA. 2001. Domain organization at the centromere and neocentromere. Dev. Cell 1:165–77 Festenstein R, Tolaini M, Corbella P, Mamalaki C, Parrington J, Fox M, Miliou A, Jones M, Kioussis D. 1996. Locus control region function and heterochromatininduced position effect variegation. Science 27:1123–25 Milot E, Strouboulis J, Trimbom T, Wijgerde M, de Boer E, Langeveld A, Tan-Un K, Vergreer W, Yannoutsos N, Grosveld F. 1996. Heterochromatin effect on the frequency and duration of LCRmediated gene transcription. Cell 87:105– 14 Festenstein R, Sharghi-Namini S, Fox M, Roderick K, Tolaini M, Norton T, Savelieve A, Kioussis D, Singh P. 1999. Heterochromatin protein 1 modifies mammalian PEV in a dose-and chromosomalcontext-dependent manner. Nature Genet. 23:457–61 Dernburg AF, Broman KW, Fung JC, Marshall WF, Philips J, Agard DA, Sedat JW. 1996. Pertubation of nuclear architecture by long-distance chromosome interactions. Cell 85:745–59

1 Feb 2002

10:23

Annu. Rev. Immunol. 2002.20:427-462. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

462

AR

SMALE

AR152-15.tex

¥

AR152-15.SGM

LaTeX2e(2001/05/10)

P1: GSR

FISHER

94. Csink AK, Henikoff S. 1996. Genetic modification of heterochromatic association and nuclear organization in Drosophila. Nature 381:529–31 95. Brown KE, Guest SS, Smale ST, Hahm K, Merkenschlager M, Fisher AG. 1997. Association of transcriptionally silent genes with Ikaros complexes at centromeric heterochromatin. Cell 91:845– 54 96. Brown KE, Baxter J, Graf D, Merkenschlager M, Fisher AG. 1999. Dynamic repositioning of genes in the nucleus of lymphocytes preparing for cell division. Mol. Cell 3:207–17 97. Lo K, Landau NR, Smale ST. 1991. LyF-1, a transcriptional regulator that interacts with novel class of promoters for lymphocyte-specific genes. Mol. Cell Biol. 11:5229–43 98. Georgopoulos K, Moore DD, Derfler B. 1992. Ikaros, an early lymphoid-specific transcription factor and a putative mediator for T cell commitment. Science 258: 808–12 99. Cortes M, Wong E, Koipally J, Georgopoulos K. 1999. Control of lymphocyte development by the Ikaros gene family. Curr. Opin. Immunol. 11:167–71 100. Trinh LA, Ferrini R, Cobb B, Weinmann AS, Hahm K, Ernst P, Garraway

101.

102.

103.

104.

105.

IP, Merkenschlager M, Smale ST. 2001. Down-regulation of TDT transcription in CD4+CD8+ thymocytes by Ikaros proteins in direct competition with an Ets activator. Genes Dev. 15:1817–32 Sabbattini P, Lundgren M, Georgiou A, Chow C-M, Warnes G, Dillon N. 2001. Binding of Ikaros to the λ5 promoter silences transcription through a mechanism that does not require heterochromatin formation. EMBO J. 20:2812–22 Cobb BS, Morales-Alcelay S, Kleiger G, Brown K, Fisher AG, Smale ST. 2000. Targeting of Ikaros to pericentromeric heterochromatin by direct DNA binding. Genes Dev. 14:2146–60 Kim J, Sif S, Jones B, Jackson A, Koipally J, Heller E, Winady S, Viel A, Sawyer A, Ikeda T, Kingston R, Georgopoulos K. 1999. Ikaros DNA-binding proteins direct formation of chromatin remodeling complexes in lymphocytes. Immunity 10:345–55 Jacobson RH, Ladurner AG, King DS, Tjian R. 2000. Structure and function of a human TAFII250 double bromodomain module. Science 288:1422–25 Lachner M, O’Carroll Rea S, Mechtler K, Jenuwein T. 2001. Methylation of histone H3 lysine 9 creates a binding site for HP1 proteins. Nature 410:116–20

6 Feb 2002

8:19

AR

AR152-15-COLOR.tex

AR152-15-COLOR.SGM

LaTeX2e(2002/01/18)

P1: GDL

Annu. Rev. Immunol. 2002.20:427-462. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

2

Figure 3 Hypothetical order of events leading to transcription initiation from the IL-12 p40 promoter.

6 Feb 2002

8:19

AR

AR152-15-COLOR.tex

AR152-15-COLOR.SGM

LaTeX2e(2002/01/18)

P1: GDL

Annu. Rev. Immunol. 2002.20:427-462. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

3

Figure 4 Order of events leading to transcription initiation from the IFN-β promoter. Derived from Agalioti et al. (72).

P1: FRK

February 1, 2002

12:47

Annual Reviews

AR152-FM

Annual Review of Immunology Volume 20, 2002

Annu. Rev. Immunol. 2002.20:427-462. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

CONTENTS FRONTISPIECE, Charles A. Janeway, Jr. A TRIP THROUGH MY LIFE WITH AN IMMUNOLOGICAL THEME, Charles A. Janeway, Jr.

xiv 1

THE B7 FAMILY OF LIGANDS AND ITS RECEPTORS: NEW PATHWAYS FOR COSTIMULATION AND INHIBITION OF IMMUNE RESPONSES, Beatriz M. Carreno and Mary Collins

29

MAP KINASES IN THE IMMUNE RESPONSE, Chen Dong, Roger J. Davis, and Richard A. Flavell

PROSPECTS FOR VACCINE PROTECTION AGAINST HIV-1 INFECTION AND AIDS, Norman L. Letvin, Dan H. Barouch, and David C. Montefiori T CELL RESPONSE IN EXPERIMENTAL AUTOIMMUNE ENCEPHALOMYELITIS (EAE): ROLE OF SELF AND CROSS-REACTIVE ANTIGENS IN SHAPING, TUNING, AND REGULATING THE AUTOPATHOGENIC T CELL REPERTOIRE, Vijay K. Kuchroo, Ana C. Anderson, Hanspeter Waldner, Markus Munder, Estelle Bettelli, and Lindsay B. Nicholson

55 73

101

NEUROENDOCRINE REGULATION OF IMMUNITY, Jeanette I. Webster, Leonardo Tonelli, and Esther M. Sternberg

125

MOLECULAR MECHANISM OF CLASS SWITCH RECOMBINATION: LINKAGE WITH SOMATIC HYPERMUTATION, Tasuku Honjo, Kazuo Kinoshita, and Masamichi Muramatsu

165

INNATE IMMUNE RECOGNITION, Charles A. Janeway, Jr. and Ruslan Medzhitov

KIR: DIVERSE, RAPIDLY EVOLVING RECEPTORS OF INNATE AND ADAPTIVE IMMUNITY, Carlos Vilches and Peter Parham ORIGINS AND FUNCTIONS OF B-1 CELLS WITH NOTES ON THE ROLE OF CD5, Robert Berland and Henry H. Wortis E PROTEIN FUNCTION IN LYMPHOCYTE DEVELOPMENT, Melanie W. Quong, William J. Romanow, and Cornelis Murre

197 217 253 301

LYMPHOCYTE-MEDIATED CYTOTOXICITY, John H. Russell and Timothy J. Ley

SIGNAL TRANSDUCTION MEDIATED BY THE T CELL ANTIGEN x

323

P1: FRK

Annu. Rev. Immunol. 2002.20:427-462. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

February 1, 2002

12:47

Annual Reviews

AR152-FM

CONTENTS

xi

RECEPTOR: THE ROLE OF ADAPTER PROTEINS, Lawrence E. Samelson INTERACTION OF HEAT SHOCK PROTEINS WITH PEPTIDES AND ANTIGEN PRESENTING CELLS: CHAPERONING OF THE INNATE AND ADAPTIVE IMMUNE RESPONSES, Pramod Srivastava CHROMATIN STRUCTURE AND GENE REGULATION IN THE IMMUNE SYSTEM, Stephen T. Smale and Amanda G. Fisher PRODUCING NATURE’S GENE-CHIPS: THE GENERATION OF PEPTIDES FOR DISPLAY BY MHC CLASS I MOLECULES, Nilabh Shastri, Susan

371

Schwab, and Thomas Serwold

395 427

463

THE IMMUNOLOGY OF MUCOSAL MODELS OF INFLAMMATION, Warren Strober, Ivan J. Fuss, and Richard S. Blumberg T CELL MEMORY, Jonathan Sprent and Charles D. Surh

GENETIC DISSECTION OF IMMUNITY TO MYCOBACTERIA: THE HUMAN MODEL, Jean-Laurent Casanova and Laurent Abel ANTIGEN PRESENTATION AND T CELL STIMULATION BY DENDRITIC CELLS, Pierre Guermonprez, Jenny Valladeau, Laurence Zitvogel, Clotilde Th´ery, and Sebastian Amigorena

495 551 581

621

NEGATIVE REGULATION OF IMMUNORECEPTOR SIGNALING, Andr´e Veillette, Sylvain Latour, and Dominique Davidson

669

CPG MOTIFS IN BACTERIAL DNA AND THEIR IMMUNE EFFECTS, Arthur M. Krieg

PROTEIN KINASE Cθ

709 IN

T CELL ACTIVATION, Noah Isakov and Amnon

Altman

RANK-L AND RANK: T CELLS, BONE LOSS, AND MAMMALIAN EVOLUTION, Lars E. Theill, William J. Boyle, and Josef M. Penninger PHAGOCYTOSIS OF MICROBES: COMPLEXITY IN ACTION, David M. Underhill and Adrian Ozinsky

761 795 825

STRUCTURE AND FUNCTION OF NATURAL KILLER CELL RECEPTORS: MULTIPLE MOLECULAR SOLUTIONS TO SELF, NONSELF DISCRIMINATION, Kannan Natarajan, Nazzareno Dimasi, Jian Wang, Roy A. Mariuzza, and David H. Margulies

853

INDEXES Subject Index Cumulative Index of Contributing Authors, Volumes 1–20 Cumulative Index of Chapter Titles, Volumes 1–20

ERRATA An online log of corrections to Annual Review of Immunology chapters (if any, 1997 to the present) may be found at http://immunol.annualreviews.org/

887 915 925

8 Feb 2002

10:1

AR

AR152-16.tex

AR152-16.SGM

LaTeX2e(2001/05/10) P1: GKD 10.1146/annurev.immunol.20.100301.064819

Annu. Rev. Immunol. 2002. 20:463–93 DOI: 10.1146/annurev.immunol.20.100301.064819 c 2002 by Annual Reviews. All rights reserved Copyright °

PRODUCING NATURE’S GENE-CHIPS:

Annu. Rev. Immunol. 2002.20:463-493. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

The Generation of Peptides for Display by MHC Class I Molecules Nilabh Shastri, Susan Schwab, and Thomas Serwold Division of Immunology, Department of Molecular and Cell Biology, University of California, Berkeley, California 94720-3200; e-mail: [email protected], [email protected], [email protected]

Key Words antigen processing, antigen presentation, proteasome, ER proteolysis ■ Abstract Gene-chips contain thousands of nucleotide sequences that allow simultaneous analysis of the complex mixture of RNAs transcribed in cells. Like these gene-chips, major histocompatibility complex (MHC) class I molecules display a large array of peptides on the cell surface for probing by the CD8+ T cell repertoire. The peptide mixture represents fragments of most, if not all, intracellular proteins. The antigen processing machinery accomplishes the daunting task of sampling these proteins and cleaving them into the precise set of peptides displayed by MHC I molecules. It has long been believed that antigenic peptides arose as by-products of normal protein turnover. Recent evidence, however, suggests that the primary source of peptides is newly synthesized proteins that arise from conventional as well as cryptic translational reading frames. It is increasingly clear that for many peptides the C-terminus is generated in the cytoplasm, and N-terminal trimming occurs in the endoplasmic reticulum in an MHC I–dependent manner. Nature’s gene-chips are thus both parsimonious and elegant.

THE PEPTIDE/MHC CLASS I DISPLAY Immune surveillance by CD8+ T cells is one of the key mechanisms for detecting and eliminating abnormal cells, including those infected with viruses or bacteria and tumor cells. CD8+ T cells probe the repertoire of peptide/MHC class I (p/MHC I) complexes on the target cell surface for novel peptides that indicate expression of foreign or abnormal gene products. Because cells cannot distinguish their normal proteins from nonself or mutant proteins, they constitutively display peptides derived from all proteins. For effective immune surveillance it is essential that the major histocompatibility complex (MHC) class I (MHC I) molecules display as large a peptide repertoire as possible to include those originating from the novel genes. Analogous to gene-chips, which contain thousands of nucleotide sequences to allow simultaneous detection of a large number of, if not all, transcripts in a 0732-0582/02/0407-0463$14.00

463

4 Feb 2002

11:41

Annu. Rev. Immunol. 2002.20:463-493. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

464

AR

AR152-16.tex

SHASTRI

¥

AR152-16.SGM

SCHWAB

¥

LaTeX2e(2001/05/10)

P1: GKD

SERWOLD

cell, the peptides displayed by MHC molecules represent the entire ensemble of polypeptides expressed within a cell. The MHC molecules and the sets of peptides they display can therefore be considered nature’s gene-chips that are probed by the CD8+ T cell repertoire. The essence of the MHC I antigen processing pathway is to load MHC I molecules with thousands of different peptides for display on the cell surface. This goal is accomplished by merging two distinct pathways: one for generating peptide receptive MHC I molecules in the endoplasmic reticulum (ER) and another for generating the pool of antigenic peptides in the cytoplasm. The two pathways merge when the cytoplasmic peptides are translocated into the ER by the transporter associated with antigen processing (TAP) and are made available to peptidereceptive MHC I molecules. The pathway that generates the peptide-receptive MHC I involves several steps and key components. MHC I molecules are associated with TAP, and tapasin acts as a bridge between the two molecules. Also present in the TAP/tapasin/MHC complex are the chaperones calreticulin and calnexin, as well as the thiol oxidoreductase ERp57. This configuration places the peptide receptive MHC I molecules in the immediate vicinity of transported peptides and also retains empty MHC I molecules until they are loaded with appropriate peptides. Tapasin may also play a role in editing and/or loading the peptides. After MHC I molecules are loaded with the peptides, they are released from this complex and exit through check-points in the ER and make their way out to the cell surface. The loading complex, and the roles of the individual components in this process, have been the subject of several excellent reviews and are not considered here (1–4). We have chosen to focus on the less well understood aspects of the pathway that generates antigenic peptides. The peptides are enormously diverse and yet precisely tailored to fit the highly polymorphic MHC I molecules present in the cell. Furthermore, the topological separation between the cytoplasm, the major site of proteolysis, and the ER, the site of MHC loading, poses a daunting challenge to the antigen processing machinery. We begin with a brief description of the strengths and limitations of the methods most commonly used to dissect the antigen processing pathway. Next, we start at the beginning of the pathway to address key unanswered questions about the source of antigenic peptides (Figure 1). Are these precursors proteins undergoing normal turnover, or are they newly synthesized gene products? We limit ourselves to endogenously synthesized precursors, leaving the question of crosspriming, in which MHC I molecules are loaded with peptides derived from exogneous antigens, to other excellent reviews (5–8). Third, we summarize our understanding of the endpoint of the pathway: the peptide/MHC complexes on the cell surface. Fourth, we examine the intermediate steps in the pathway, following the precursors as they are transformed from polypeptide chains into the precise peptides that are presented by the MHC I molecules. We focus on the proteolytic events required to generate the peptides. Where and how are the precursors degraded to generate the antigenic peptides? Is the precisely cleaved peptide

4 Feb 2002

11:41

AR

AR152-16.tex

AR152-16.SGM

LaTeX2e(2001/05/10)

P1: GKD

Annu. Rev. Immunol. 2002.20:463-493. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

LOADING NATURE’S GENE-CHIPS

465

Figure 1 Overview of the MHC I antigen processing pathway. See text for details. Question marks indicate key outstanding issues.

product generated during antigen processing in the cytoplasm, or does the cytoplasmic product contain additional flanking residues that require trimming in the ER? What is the nature of the proteases involved and what are their proteolytic products? And finally, what is the role of the MHC I molecules in antigen processing?

METHODS FOR STUDYING THE MHC I ANTIGEN PROCESSING PATHWAY Processing of polypeptides in the MHC I pathway results in surface expression of the p/MHC I complex, whose raison d’ eˆ tre is to serve as a potential ligand for CD8+ T cells. T cell assays have therefore remained the most convenient method to detect the existence of a specific p/MHC complex in intact antigen presenting cells (APC). Conventional assays measure the activation of cytotoxic CD8+ T cells (CTL) by specific lysis of APC or by cytokine secretion. The presence of the p/MHC can also be assayed using T cell hybridomas that secrete IL-2 (9) or accumulate β-galactosidase (lacZ) (10, 11). These assays are often used to compare p/MHC levels on various APC. However, such comparisons can be difficult because even low copy numbers can often maximally activate T cells. Many CD8+ T cells can detect peptides at low picomolar concentrations; nanomolar concentrations are generally more than

4 Feb 2002

11:41

Annu. Rev. Immunol. 2002.20:463-493. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

466

AR

AR152-16.tex

SHASTRI

¥

AR152-16.SGM

SCHWAB

¥

LaTeX2e(2001/05/10)

P1: GKD

SERWOLD

saturating. In this case increasing p/MHC copy number will not increase the activation readout. Two apparently equivalent APC may in fact have very different p/MHC levels. A technically demanding but more informative method for measuring processed peptides is to fractionate cell extracts by high performance liquid chromatography (HPLC) and to detect the naturally processed peptides in an exogenous assay using appropriate APC and T cells (12, 13). The high resolution of HPLC permits the identification of extracted peptides as well as their relative amounts by comparison with synthetic peptide standards. MHC–bound peptides have also been detected and identified by massspectrometry (14, 15). This technique has been particularly successful in revealing posttranslational modifications in the antigenic peptides (16–18). Another method to detect the p/MHC I complex in the APC using a monoclonal antibody has recently emerged. This antibody was generated in the Germain laboratory, and it recognizes the ovalbumin (OVA)-derived SIINFEKL (SL8) peptide bound to the mouse Kb MHC I molecule (19). A final method measures the total p/MHC levels on the cell surface using antibodies specific for peptide loaded MHC molecules (20). This provides a window on the overall efficiency of the antigen processing machinery without requiring knowledge of specific peptides or the T cells to detect them (21–23). With assays to measure the final p/MHC, the study of antigen processing involves experimentally manipulating the precursors and the cells in various ways and correlating changes in the p/MHC expression with the intervening steps.

ANTIGENIC PEPTIDES ARE DERIVED FROM A WIDE RANGE OF PRECURSORS With the notable exceptions of the Epstein Barr virus–derived EBNA I protein and the cytomegalovirus (CMV)-derived 72K principal immediate-early protein, no endogenously synthesized precursor appears to be excluded from entry into the antigen processing pathway (24–26). Rammensee et al. have compiled a comprehensive list of hundreds of self and nonself peptides known to be presented by MHC I molecules (27). This list includes peptides derived from proteins with widely varying functions, cellular locations, and abundance. This impressive diversity is important and suggests that omissions, which would constitute a blind spot for the immune system, are rare; the antigen processing mechanism operates on most, if not all, proteins available in the cell. The original list of peptides bound to human MHC I molecules also included many peptides of unknown origin. Using the recently completed draft of the human genome sequence, we have identified putative precursors of some of these peptides (Table 1). Even this short list of 21 precursors contains a strikingly broad spectrum of proteins.

4 Feb 2002

11:41

AR

AR152-16.tex

AR152-16.SGM

LaTeX2e(2001/05/10)

P1: GKD

467

LOADING NATURE’S GENE-CHIPS TABLE 1 Potential sources for unknown MHC binding peptidesa

Annu. Rev. Immunol. 2002.20:463-493. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

MHC ligand

MHC molecule

Protein/gene

GenBank accession number/celera protein number (hCP)

AA position

APRQPGLMA

HLA-B∗ 5501

16.7-kD protein

NP 057223

49–57

HPKYKTEL

HLA-B8

Butyrate response factor 2; TIS11B protein; zinc finger protein homologous to Zfp-36 in mouse

NP 008818; Q07352; NP 003398

189–196; 150–157; 139–146

DRHERITKL

HLA-B14

Chromosome 11 open reading frame 2

NP 037397

159–167 125–133

SIRDGVRAY

HLA-B∗ 4601

Crystallin, zeta-like 1

NP 005102

KIKSFEVVF

HLA-A3

Dihydroxyvitamin D3induced protein

1090504

TYYGSFVTR

HLA-A∗ 3302

Eukaryotic translation initiation factor 3, subunit 3

NP 003747

123–131

STYYGSFVTR

HLA-A∗ 1101

Eukaryotic translation initiation factor 3, subunit 3

NP 003747

122–131

SQFGGGSQY

HLA-B∗ 1501

Eukaryotic translation initiation factor 3, subunit 7

NP 003744

61–69

FIKDGSSTY

HLA-B∗ 4601

hCP37886

hCP37886.1

489–497

ALSNLEVKL

HLA-A∗ 0201

hCP38933

hCP38933.1

155–163

EHAGVISVL

HLA-B∗ 3801

Hepatitis B virus xinteracting protein

NP 006393

40–48 443–451

6–14

KRFEGLTQR

HLA-B∗ 2705

KIAA0965 protein

BAA76809

RRFTRPEH

HLA-B∗ 2705

KIAA1063 protein

BAA83015

435–442

EVAPPEYHR

HLA-A∗ 6801

MORF-related gene 15

NP 006782

312–320

EVAPPEYHRK

HLA-A∗ 6801

MORF-related gene 15

NP 006782

312–321

RRISGVDRY

HLA-B∗ 2705

NADH dehydrogenase 1, alpha subcomplex 1

NP 004532

52–60

HLPETKFSEL

HLA-Cw∗ 0102

Non-lens beta gammacrystallin-like protein

AAB53791

856–865

DAYALNHTL

HLA-B∗ 5101

POU domain class 2, associating factor 1 Ribosomal protein L15

Q16633

243–251

EVILIDPFHK

HLA-A∗ 6801

NP 002939

131–140

IAPTGHSL

HLA-Cw∗ 0102

Septin 6; hypothetical protein FLJ10849

Q14141; NP 060713

152–159; 154–161

AYVHMVTHF

HLA-A24

Testis enhanced gene transcript protein

AAB87479

45–53

a

The peptide sequences are from the list of sequences marked as “unknown origin” in Reference (27). Potential sources were identified by searching the Genbank and Celera human genome databases. The AA position indicates where in the open reading frame the peptide sequence is located.

4 Feb 2002

11:41

468

AR

AR152-16.tex

SHASTRI

¥

AR152-16.SGM

SCHWAB

¥

LaTeX2e(2001/05/10)

P1: GKD

SERWOLD

Annu. Rev. Immunol. 2002.20:463-493. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

ARE THE PRECURSORS OLD OR NEWLY SYNTHESIZED PROTEINS? Antigenic peptides clearly represent a wide range of gene products. The nature of the precursors, however, remains an intriguing and open question. A key issue is, when does a protein enter the antigen processing pathway? One can envision four stages in the life of an endogenous protein: The polypeptide is translated by the ribosomes, the polypeptide folds, it carries out its biological function, and it is subsequently degraded. For antigen processing, using only those proteins that are being turned over after their biological function has been completed is economical and politically correct. Why not re-use parts of a protein at the end of its useful life? Contrary to this idea, recent data indicate that the bulk of antigenic peptides come from proteins degraded immediately after synthesis. Obtaining antigenic peptides from newly synthesized proteins alleviates the theoretically difficult requirement that proteins or their fragments be retrieved from different cellular compartments such as the nucleus for entry into the antigen processing pathway. Furthermore, because proteins are sampled concomitantly with their synthesis, this provides for early detection of novel peptides derived from intracellular pathogens or mutations. If polypeptides are marked for rapid degradation because they failed to fold properly or were incorrectly translated to begin with, obtaining antigenic peptides from these precursors retains the same advantage of economy as using protein turnover. Historically it was generally believed that antigenic peptides were derived from the turnover of mature proteins. In support of this model, mature proteins artificially introduced into the cell (28) or secreted into the cytoplasm by the intracellular bacterium Listeria monocytogenes (29) were processed into antigenic peptides. Additionally, proteins can be retrieved from several subcellular compartments for entry into the pathway, indicating that there is no obligate link between protein synthesis and peptide generation. The nonclassical MHC I molecule H2-M3 presents N-formylated peptides derived from proteins synthesized in the mitochondria (30, 31). In two instances peptides with Asn to Asp modifications were found; this is consistent with glycosylation in the ER followed by retrieval into the cytosol for entry into the antigen processing pathway (17, 32). Several additional cases of peptides with posttranslational modifications have been reported (16, 18). Furthermore, as predicted, when exogenous proteins were introduced into cells, the rate of peptide generation mirrored the protein half-life. Two studies demonstrating this relationship compared peptide generation from stable and unstable proteins that were constructed taking advantage of the N-end rule: Proteins with certain N-terminal amino acids are rapidly degraded by the 26S proteasome in a ubiquitin-dependent manner (33–35). These studies unequivocally demonstrated that proteasome-mediated turnover of mature proteins can produce antigenic peptides, but they did not address the question of whether this turnover contributes a significant supply of peptides derived from endogenously synthesized precursors.

4 Feb 2002

11:41

AR

AR152-16.tex

AR152-16.SGM

LaTeX2e(2001/05/10)

P1: GKD

Annu. Rev. Immunol. 2002.20:463-493. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

LOADING NATURE’S GENE-CHIPS

469

One prediction of the model in which newly synthesized proteins are the main entrants into the antigen processing pathway is that the time course of peptide generation from an endogenous precursor will correlate with its synthesis rather than degradation of the mature protein. In the first study of an endogenous precursor, Townsend and colleagues compared peptide generation from two vaccinia constructs that encoded the influenza nucleoprotein (NP) protein and differed only by the N-terminal amino acid (36). Using CTL lysis as a measure of p/MHC expression, they found that cells expressing the unstable NP protein were more sensitive to killing by NP/Db-specific CTL than those expressing the stable protein. However, lysis by an NP/Kk-specific CTL did not correlate with the difference in stability of the same precursors. In a similar experiment, cells were infected with vaccinia encoding two forms of the influenza virus matrix protein (37). Cells expressing the stable or unstable proteins were equally sensitive to CTL lysis. One difficulty with both studies is that CTL lysis is not a quantitative measure of peptide abundance; it is possible that in the cases in which killing was equivalent, peptide abundance may have differed, but the CTL response was maximal even at the lower peptide amount. A third study from our laboratory found no difference in the amounts of naturally processed peptides generated in cells expressing pairs of DNA constructs that differed in the N-terminus of the encoded proteins (38). Although the precursor half-lives differed by six- to ninefold, the amounts of naturally processed peptides extracted from transfected cells were comparable. Unlike the CTL lysis assays, this experiment was quantitative; a difficulty was that peptides were detectable only after both proteins had gone through multiple halflives. Nevertheless, these studies questioned the notion that peptides are generated from mature protein turnover, and they suggested a link between protein synthesis and antigen processing. Reits et al. took an entirely different and elegant approach to the study of antigen processing and found a striking correlation between protein synthesis and peptide supply. They monitored the availability of peptides for MHC I presentation by measuring TAP activity in living cells, using the fluorescence bleaching/recovery technique (39). TAP moves peptides from the cytoplasm into the ER, a key stage in the journey of more than 90% of antigenic peptides. After the addition of cycloheximide, a protein-synthesis inhibitor, TAP activity rapidly ceased, within 25 min in uninfected and 40 min in influenza-infected cells, suggesting a paucity of peptides. The simplest interpretation of these results is that the majority of antigenic peptides comes from newly synthesized proteins. Schubert et al. took another novel approach (40). They argued that if antigenic peptides came primarily from newly synthesized proteins, then inhibiting the proteasome would cause these proteins to accumulate quickly. They treated cells with and without a proteasome inhibitor, pulse-labeled them with [35S]methionine, and measured the total amount of radioactivity in fractionated proteins larger than 14 kDa. Addition of the proteasome inhibitor increased the total amount of radioactivity recovered immediately after labeling, and the increase occurred preferentially in high molecular weight proteins, a fraction of which were ubiqitinated. They

4 Feb 2002

11:41

Annu. Rev. Immunol. 2002.20:463-493. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

470

AR

AR152-16.tex

SHASTRI

¥

AR152-16.SGM

SCHWAB

¥

LaTeX2e(2001/05/10)

P1: GKD

SERWOLD

calculated that 30% of newly synthesized proteins were degraded within 10 min. They referred to these proteins as defective ribosomal products (DRiPs) and suggested that they represent a major source of processed peptides. Together these studies indicate that newly synthesized proteins are a major source of precursors for antigen processing. In fact, they raise the question of whether proteins at the end of their natural lives enter the antigen processing pathway at any significant rate. Nevertheless, the links between protein synthesis and protein degradation remain to be disentangled. Notably, both of these studies relied heavily on the use of inhibitors of central cellular processes, and secondary effects of these inhibitors were not ruled out. If antigenic peptides are in fact derived primarily from newly synthesized proteins, the question arises of which new proteins are targeted for the processing pathway. Proteins that are defective in some way would naturally be degraded rapidly. Yewdell et al. proposed that DRiPs consist of prematurely terminated polypeptides and misfolded polypeptides produced from translation of bona fide mRNAs in the proper reading frame (41). Experiments have yet to verify the importance of misfolding and premature termination in generating antigenic peptides. C-terminally truncated polypeptides are probably not a major source of antigenic peptides because an N-terminal preference is not evident in the list of known peptides compiled by Rammensee (27) or in the new list (Table 1).

CRYPTIC TRANSLATION AS A SOURCE OF ANTIGENIC PEPTIDES In addition to misfolded and prematurely terminated proteins, it is increasingly clear that antigenic peptides can come from sources other than translation of the primary open reading frame. Evidence comes from several experiments in which a T cell clone was used to define its cognate p/MHC I complex (Table 2). In one case, a murine leukemia, the antigenic peptide was tracked to the 50 untranslated region of an mRNA transcript. In nine instances—six human cancers and three virus-infected mouse models—antigenic peptides were traced to alternative open reading frames of mRNAs encoding longer proteins. In four cases of human melanomas, antigenic peptides were tracked to introns or to the intron-exon junction. These examples indicate that unconventional peptides can elicit cytotoxic T lymphocyte responses. We suspect that such peptides may in fact play an important role in immune surveillance. Fundamentally the immune system cares about only one attribute of a peptide: whether it is self or nonself. Whether the peptide came from a functional protein or an unconventional source is irrelevant. DNA contains a wealth of information that is not captured by translation of the primary open reading frame; translation of alternate reading frames, intronic sequences, and 50 and 30 “untranslated” regions of mRNA can make this information available. The resulting polypeptides could be particularly valuable in cases in which the

4 Feb 2002

11:41

AR

AR152-16.tex

AR152-16.SGM

LaTeX2e(2001/05/10)

P1: GKD

LOADING NATURE’S GENE-CHIPS

471

Annu. Rev. Immunol. 2002.20:463-493. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

TABLE 2 Examples of cryptic translation Cell type

Peptide source

Reference

Stably transfected murine cell line (Lmtk−)

Region downstream of stop codon introduced into the SV40 gene

(115)

Stably transfected murine tumor cell line (P1.HTR tk−)

Promoterless fragment of the P91A gene

(116)

Murine cell lines (BALB/3T3, L929) infected with MMLVderived retrovirus

Region downstream of stop codon introduced into the NP gene

(117)

Murine cell line (P815) infected with vaccinia

Region downstream of stop codon introduced into the HA gene

(118)

Transiently transfected simian cell line (COS)

Region of the ovalbumin gene placed out of frame with the primary AUG, region of the NP gene placed out of frame with the primary AUG. Neither gene had an alternative AUG start.

(119)

BALB C radiationinduced leukemia

50 untranslated region of c-act oncogene

(120)

Human melanoma

Region of the MUM-1 gene spanning intronexon junction

(121)

Transiently transfected simian cell line (COS), stably transfected murine cell line (Hepa1)

Region of gene out-of-frame with the primary AUG, and without an alternative AUG start

(122)

Transiently transfected simian cell line (COS), stably transfected murine cell line (Lmtk−)

Region of gene without upstream AUG

(48)

Cells from vacciniainfected mouse

Region downstream of stop codon introduced into the NP gene. Mice infected with vaccinia encoding the mutant NP generated CTL against downstream epitopes.

(47)

Cells from vacciniainfected mouse

Region downstream of stop codon introduced into the NP gene. Mice infected with vaccinia encoding the mutant NP generated CTL against downstream epitopes.

(123)

Human melanoma

Intron of the N-acetyl glucosaminyltransferase V gene

(124) (Continued)

4 Feb 2002

11:41

472

AR

AR152-16.tex

SHASTRI

¥

AR152-16.SGM

SCHWAB

¥

LaTeX2e(2001/05/10)

P1: GKD

SERWOLD

Annu. Rev. Immunol. 2002.20:463-493. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

TABLE 2 (Continued) Cell type

Peptide source

Reference

Human melanoma, normal cultured melanocyte cell line

Alternate ORF of gp75

(125)

Murine cell line (L929) infected with vaccinia

Alternate ORF of mutant influenza NP. The authors found that out-of-frame epitopes were expressed even when the AUG for the primary ORF was in an excellent context for translation initiation. The level of expression of out-of-frame epitopes increased as the context of the primary AUG worsened.

(126)

Human melanoma, normal cultured melanocyte cell lines

Intron of gp100 gene

(127)

Human melanoma

Intron of TRP-2 gene

(128)

Cells from mouse infected with LP-BM5 MAIDS retroviral complex

Alternate ORF of MAIDS gag gene. Mice infected with MAIDS generated CTL against this epitope.

(129)

Human squamous cell carcinoma, lung adenocarcinoma

Alternate ORF of SARC-1

(130)

Human melanoma, breast cancer cells

Alternate ORF of NY-ESO-1

(131)

Human melanoma

Alternate ORF of CAMEL

(132)

Human B cell acute lymphoblastic leukemia, Epstein-Barr virus–transformed B cells

HB-1, initiated at CUG codon

(133)

Transiently transfected simian cell line (COS)

A minigene that was (a) initiated with a nonAUG codon (CUG) and (b) 30 of a stop codon terminating another minigene in a conventional translational context. The CUG initiation codon was translated as leucine rather than the canonical methionine.

(49)

Renal cell carcinoma

Alternate ORF of the intestinal carboxyl esterase gene, initiated at ACG

(134)

Human renal cell carcinoma, normal kidney and liver cells

Alternate ORF of the M-CSF gene

(135)

4 Feb 2002

11:41

AR

AR152-16.tex

AR152-16.SGM

LaTeX2e(2001/05/10)

P1: GKD

Annu. Rev. Immunol. 2002.20:463-493. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

LOADING NATURE’S GENE-CHIPS

473

conventional protein had few epitopes that could squeeze through the bottlenecks in the antigen processing pathway. These bottlenecks include the ability of cellular proteases to produce the processed peptide, the ability of TAP to transport the peptide into the ER, and the ability of the peptide to bind an available MHC I molecule (42–44). Furthermore, the precursors of these peptides, which presumably have no function in the cell, are likely to be shuttled immediately to the antigen processing pathway. They would thus be early sentinels of an infection or mutation and perfect candidates for the newly synthesized proteins that are degraded by the proteasome and transported by TAP as discussed above. These peptides are often termed cryptic, because the mechanism of their translation is unknown. Thierry Boon and Aline Van Pel were the first to offer an explanation in the form of the pepton hypothesis (45). They argued that antigenic peptides were not degradation products of cellular proteins at all but instead were generated directly by the “autonomous transcription and translation of short subgenic regions” called “peptons” using a novel RNA polymerase. Although there is no evidence to date for this specialized transcriptional machinery, Boon & Van Pel deserve credit for directing attention toward unconventional sources of antigenic peptides. A second mechanism that could account for these peptides is mis-splicing of primary transcripts. Mis-splicing could explain translation of intronic sequences; it could also lead to transcripts missing the primary AUG codon, allowing translation to begin at an out-of-frame initiation codon. In this context a recent report of coupled transcription and translation in the nucleus is intriguing; the authors suggest that this accounts for 10–15% of protein synthesis in the cells and provides a mechanism for nonsense-mediated decay (46). The proximity of the proteasome to the nuclear translational site could make these polypeptides available to the antigen processing pathway as well. Cryptic peptides could also be generated by the ribosome translating a conventional mRNA transcript in an unconventional way. The accepted model of translation is that the ribosome binds mRNA at the 50 cap, scans in the 30 direction for the first AUG, and translates nucleotides in triplets until a stop codon is reached. Cryptic peptides would be generated if the ribosome (a) bound the mRNA at the 50 cap but scanned through the primary AUG to start at a downstream initiation codon, (b) bound the mRNA at an internal site and initiated translation at a codon downstream of the primary AUG, or (c) initiated translation at the primary AUG and frameshifted during elongation. Bullock & Eisenlohr demonstrated that ribosomal scan-through of the primary AUG is a potent mechanism for translation of alternate reading frames (47). The pool of potential precursors was widened by studies showing that initiation of antigenic peptides can occur not only at alternate AUGs but at several additional codons (48, 49). Malarkannan also implicated a novel initiation mechanism for generation of antigenic peptides: When initiation occurred at the CUG codon (one of the non-AUG codons), it was decoded as leucine rather than the canonical methionine. The underlying molecular

4 Feb 2002

11:41

474

AR

AR152-16.tex

SHASTRI

¥

AR152-16.SGM

SCHWAB

¥

LaTeX2e(2001/05/10)

P1: GKD

SERWOLD

Annu. Rev. Immunol. 2002.20:463-493. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

mechanisms that allow the translation of these polypeptides and target them into the antigen processing pathway are unknown. Cryptic peptides could enhance immune surveillance by adding to the diversity of peptides displayed on the cell surface. However, many open questions remain; the most basic issue requiring resolution is whether cryptic translation is a rare anomaly, seen only in virally infected, transformed, or transfected cells, or whether it occurs systematically in normal cells as well. A second important question is the relative contribution of cryptic translation to the total peptide pool, in terms of numerical abundance, diversity, and immunological significance.

THE ANATOMY, DIVERSITY, AND ABUNDANCE OF NATURALLY PROCESSED PEPTIDES Whereas our knowledge of the pool of precursors for antigen processing is still rather murky, the picture of the final product is crystal clear. The final products are short, usually 8–10 residue, peptides bound to MHC I molecules on the cell surface. A cell from a heterozygous individual expresses several hundred thousand copies of up to six different MHC I molecules. Each MHC I molecule contains a single peptide. Thus, it is possible for a single cell to express thousands of distinct p/MHC I complexes. Most peptides displayed by a given MHC I molecule share a consensus motif defined strictly by size and the presence of conserved residues at defined positions. The consensus motifs for MHC I–bound peptides were first discovered by sequencing the mixture of peptides eluted from purified MHC I molecules (44, 50). These “simple” yet elegant experiments yielded remarkably profound insights into the mechanism that permits MHC I to bind a diverse set of peptides that is essential for effective immunesurveillance. As expected, the MHC I–bound peptide pool was heterogenous, but conserved amino acids were clearly present at one or two internal, usually the p2, p3, or p5, positions and at the C-terminus. In contrast, residues at other positions were highly variable. The variability of these six or seven residues in octamer or nonamer peptides allows up to 206–7 (∼1 × 108) different peptides to be presented by any one MHC I molecule. The crystal structures of many peptide/MHC I complexes subsequently showed that the peptides were primarily tethered to the MHC molecule by their N- and C-termini, the side chains of the conserved carboxyl and internal residues, and the peptide backbone (51). Notable exceptions to the peptide length, consensus motifs, and mode of binding have been found among the peptides presented by MHC I molecules. For example, the murine Kb MHC I, which usually presents octamers with the consensus sequence XXXX[F,Y]XX[I,L,M] (amino acid residues are in single letter code; X = any amino acid), can accommodate a nonamer peptide, FAPGNYPAL, with a central bulge (52). Likewise the human HLA-A2, which usually presents X[L]XXXXXX[L,V] nonapeptides, can also bind to a MLLSVPLLLG decamer with the C-terminal glycine residue extending out of the MHC I antigen-binding

4 Feb 2002

11:41

AR

AR152-16.tex

AR152-16.SGM

LaTeX2e(2001/05/10)

P1: GKD

Annu. Rev. Immunol. 2002.20:463-493. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

LOADING NATURE’S GENE-CHIPS

475

groove (53). In general these and other exceptions are a result of amino acid combinations in the antigenic peptide that alter the peptide conformation or can serve as secondary anchor residues to provide requisite stability to the p/MHC I complex (54). Together the variable residues of the peptide as well as the surface of the MHC molecule interact with the TCR to provide highly specific recognition of even very subtle changes in the antigenic peptides (55). The number of peptides that can be presented by a given MHC I molecule is constrained by the consensus motifs required for binding. Even this constraint, however, is alleviated because a heterozygous individual expresses up to six different MHC I molecules; as of July 2001, almost 800 MHC I alleles had been discovered in the human population as a whole (http://www3.ebi.ac.uk/Services/imgt/hla/cgibin/statistics.cgi). The highly polymorphic nature of the MHC I loci generates amino acid substitutions within the antigen-binding groove of the MHC molecules. As a result the complementary pockets of the MHC I that accommodate the primary and secondary anchor residues of the antigenic peptide vary extensively. Accordingly, the consensus motifs are remarkably variable among the peptides bound to different MHC I molecules. This is illustrated by the presence of over 48 different consensus motifs among the 69 MHC I molecules surveyed by Rammensee and his colleagues (27). The “X-Proline(P)-Xn” motif represents a notable exception to the high variability found among the MHC-bound peptide motifs. The “X-P-Xn” motif accounts for over 20% of all known motifs in the human and other species (27, 56). However, the predominance of this motif is paradoxical because TAP excludes peptides with this motif from transport into the ER (43, 57). We discuss a possible explanation for this paradox later. Nevertheless, the high overall variability of consensus motifs makes it possible that any one individual can present six entirely distinct sets of peptides. This peptide diversity provides the basis for efficient immunesurveillance by CD8+ T cells but also poses special challenges for the antigen processing mechanisms responsible for generating these peptides for loading the MHC I molecules. The diversity in the sets of peptides presented by the MHC I molecules also implies that the overall abundance of individual peptide/MHC I complexes detected by antigen-specific CD8+ T cells must be quite low. CD8+ T cells possess an impressive capacity to recognize and respond to as few as one p/MHC on the target cell surface (58). Compelling evidence directly demonstrating the complexity of the peptide pool at the level of individual peptides has been obtained using sensitive mass spectrometric analysis (59). The authors estimated that although some peptides were presented at several hundred copies/cell, others, possibly thousands, could be presented at far lower levels. Most important, the low abundance of specific p/MHC complexes displayed on the cell surface has been confirmed in experiments in which the naturally processed peptides were quantitated in antigen presenting cell (APC) extracts and were in many cases present at fewer than 10 copies/cell (60–62). The antigen processing machinery faces an extraordinary challenge: It must produce a set of peptides that are enormously diverse and yet are precisely cleaved

4 Feb 2002

11:41

476

AR

AR152-16.tex

SHASTRI

¥

AR152-16.SGM

SCHWAB

¥

LaTeX2e(2001/05/10)

P1: GKD

SERWOLD

to conform to the consensus motifs of the MHC I molecules expressed in the cell. The remainder of this review addresses some of the key questions about how this is achieved.

Annu. Rev. Immunol. 2002.20:463-493. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

THE RAMMENSEE PARADOX: MHC I MOLECULES INFLUENCE ANTIGEN PROCESSING An absolutely critical issue in understanding the transformation of precursor polypeptides to the exactly cleaved peptides is the role of the MHC I molecules. Are MHC I molecules passive recipients of the peptides made available to them, or do they affect the antigen processing events? The important role of MHC I molecules in influencing the outcome of the antigen processing reactions was first underscored a dozen years ago by the Rammensee laboratory. Soon after the recognition that naturally processed peptides could be detected in cell extracts fractioned by high performance liquid chromatography (HPLC) (12, 13, 63), Rammensee and his colleagues compared cells expressing the antigen and the appropriate MHC I molecule with cells that expressed the antigen alone (64, 65). Most surprisingly, they found that each of three minor histocompatibility and two influenza virus–specific CTLs could detect their cognate peptides only in extracts of cells that expressed the appropriate MHC I molecule. With rare exceptions (66), these observations of the ability to detect the final processed peptide only in the presence of the appropriate MHC I have been consistently reproducible in different model systems and laboratories (67–71). How the MHC I molecules in the ER could influence the naturally processed peptide pool believed to be generated in the cytoplasm has remained an enduring paradox with profound implications for understanding the antigen processing pathway. To explain the Rammensee paradox, one hypothesis, termed here the “protection” model, was advanced by Elliott and colleagues (72). The protection model suggested that the processed peptides were generated in the cytoplasm but were extraordinarily labile and rapidly degraded. The processed peptides were therefore not detectable in cell extracts unless they were protected by binding to MHC I. An alternative hypothesis proposed by Falk et al. suggested that the MHC I molecules themselves were essential for generating the final version of the naturally processed peptides by acting as “templates” (64). Most importantly these two hypotheses had very different implications for the nature of the cytoplasmic peptide pool. The protection model implied that cytoplasmic proteolysis generated the final processed peptide, which was rapidly transported by TAP and immediately loaded onto the MHC I molecules in the ER (72). However, the “template” model did not require that the final peptide be generated in the cytoplasm or even be labile. The cytoplasmic precursors could have additional flanking residues, but ER proteases would be essential for trimming the extra flanking residues under the influence of the MHC I molecules (64).

4 Feb 2002

11:41

AR

AR152-16.tex

AR152-16.SGM

LaTeX2e(2001/05/10)

P1: GKD

LOADING NATURE’S GENE-CHIPS

477

Recent studies discussed below provide plausible explanations for the resolution of this paradox and insights into the intermediate steps between the precursors and their final peptide products.

Annu. Rev. Immunol. 2002.20:463-493. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

MAKING THE CUTS Regardless of the source and the sampling criteria used, all normal precursors must be fragmented to generate the precisely cleaved peptides presented by MHC I molecules. Whereas passage through the proteolytic steps is not obligatory— antigenic precursors that exactly match the final peptide product are efficiently presented by MHC I (49, 70)—all known naturally processed peptides are embedded within a polypeptide chain. This internal location dictates that precise N- and C-terminal cuts must be made to remove all flanking residues. It is conceivable that a single endopeptidase, such as the proteasome, could cleave the antigenic precursors in a concerted manner to release the final peptide and satisfy the expectation of the protection model (73). However, the recent discovery of proteolytic intermediates and related evidence indicate that different proteolytic mechanisms account for generating the precise C- and N-termini of the final peptide products. Proteolytic intermediates longer than the final peptide cannot normally be detected because they are inactive in exogenous T cell assays. This was a major limitation that prevented analysis of such intermediates earlier (64, 69). Paz et al. devised a new method that combined HPLC fractionation of cell extracts to separate the different intermediates with protease treatments to release the optimally active antigenic peptides from inactive precursors, allowing their detection by T cell activation (74, 75). This new method for analysis of processed peptides revealed that cells do not generate the exact peptide in a single step. Instead the cells generated a number of proteolytic intermediates containing the antigenic peptide with one to four or more flanking residues. The composition of intermediates in the cytosol was distinct from that of the ER, and only the shorter set of peptides was transported by TAP. The Paz analysis provides a key insight into the Rammensee paradox. Antigen processing in the cytoplasm did not generate exactly cleaved peptides as envisaged in the protection model (72). Next we discuss our understanding of the proteolytic mechanisms that generate the final peptides with the exact “start,” the N-terminus, and the “end,” the Cterminus.

STARTING AT THE “END”: GENERATING THE C-TERMINUS OF ANTIGENIC PEPTIDES The Paz study discussed in the previous section was limited by the precursor design in which the antigenic peptide was located at the C-terminus (74). Unlike natural precursors, generation of the intermediates that yielded the final SL8 peptide from

4 Feb 2002

11:41

478

AR

AR152-16.tex

SHASTRI

¥

AR152-16.SGM

SCHWAB

¥

LaTeX2e(2001/05/10)

P1: GKD

SERWOLD

this precursor required only N-terminal and no C-terminal cleavages. How do cells generate the precise ends, the C-termini of antigenic peptides? Several lines of evidence indicate that the C-terminal ends of antigenic peptides are generated in the cytoplasm and are made by the proteasome.

Annu. Rev. Immunol. 2002.20:463-493. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

PROTEASOMES IN THE ANTIGEN PROCESSING PATHWAY Antigen processing begins in the cytoplasm, and the proteasome plays a central role. Following the clue that two MHC linked genes, LMP2 and LMP7, were subunits of the multicatalytic proteasomes (76), the Rock and Goldberg laboratories provided the first key insight into the proteolytic mechanism for generating antigenic peptides (77). They showed that inhibiting proteasome activity caused a profound block in the turnover of short- and long-lived intracellular proteins and also abrogated the generation of p/MHC I complexes in cells. Numerous subsequent studies have confirmed that the proteasome is the key protease in the generation of p/MHC I complexes (78). For most antigenic peptides, specifically inhibiting the proteasome causes loss of presentation, but for reasons not yet clear the presentation of some peptides actually increases in cells treated with proteasome inhibitors (21, 79, 80). Recognizing the importance of the proteasomes in generating the processed peptides, many studies have focused on defining its structure and role in antigen processing. These studies have been recently reviewed (78, 81, 82). The proteasome is a complex and heterogenous structure with several distinct catalytic activities. In addition to the constitutive 20S proteasome expressed in most cells, the “immunoproteasome” is the predominant species in lymphoid tissues. The latter includes the γ -interferon–inducible subunits LMP2, LMP7, and MECL-1 that replace the constitutive subunits (83). These 20S core particles can be further modified by the addition of a 19S cap structure as well as by the γ -interferon–inducible PA28α and PA28β. The proteasome, owing to its central role in many distinct biological processes such as signal transduction pathways and the cell cycle, is essential for cell viability. Notwithstanding reports that other proteases can replace proteasome function (84, 85), knock-outs of the delta or X subunit of the constitutive proteasome (replaced by LMP2 and LMP7, respectively, in the immunoproteasome) are lethal at the very early pre-implantation state of development (J. Monaco, personal communication). However, the immunoproteasome is not essential for viability, and its absence caused by knock-out of LMP2 (86), LMP7 (87), and the MECL-1 genes (J. Monaco, personal communication) correlates with subtle defects in antigen presentation. Likewise PA28 knockout mice are impaired in their ability to process some endogenous and exogenous antigens and to generate CTL responses (88). Note that the loss of TAP1 (89) or tapasin (90, 91) causes severe five- to tenfold reduction in the overall presentation of antigenic peptides by MHC I molecules.

4 Feb 2002

11:41

AR

AR152-16.tex

AR152-16.SGM

LaTeX2e(2001/05/10)

P1: GKD

Annu. Rev. Immunol. 2002.20:463-493. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

LOADING NATURE’S GENE-CHIPS

479

In contrast, the loss of LMP2, LMP7, MECL-1, or PA28 causes far less severe defects. It is therefore very likely that the constitutive proteasome and/or other proteases are sufficient for generating most antigenic peptides for MHC I molecules, and the immunoproteasome and other modifiers influence the quality or quantity of the processed peptide pool under some circumstances. Delineating the role of the proteasome subunits and modifications in antigen processing is an active area of research. In vitro assays using colorimetric and peptide substrates have consistently identified three major cleavage specificities of the proteasome (83). These specificities promote cleavage after hydrophobic, basic, and negatively charged amino acids. Notably, introduction of the IFN γ –inducible subunits enhances cleavages after the basic and hydrophobic residues, consistent with peptide C-terminal preferences MHC I.

PROTEASOMES MAKE THE C-TERMINAL CUT Several lines of evidence indicate that the proteasome is required to generate the C but not the N terminus of the final peptide. Craiu et al. used proteasome inhibitors to block presentation of the ovalbumin-derived SL8/Kb complex in cells expressing precursors that were extended at the N- or C-termini (92). They noticed that the inhibitors could block presentation of the SL8 peptide when it contained even one C-terminal flanking residue but not with up to 25 N-terminal flanking residues. Similar observations with two other antigenic peptides (93) suggested that proteasomal activity was essential only for generating the precise C-terminus and was not required for generating the precise N-terminus. Independently another study used the HPLC/enzymatic method to detect the SL8 peptide as well as the Cterminal extended SL8-I peptide in cells expressing Kb and a precursor with SL8-I at its C-terminus (75). Again in this study, the generation of naturally processed SL8 but not the SL8-I peptide required active proteasomes. A third study examined the generation of the SL8 peptide analogs by the constitutive and immunoproteasome in vitro (94). In a significant advance over other similar studies, the authors were able to use whole ovalbumin rather than synthetic peptides as a substrate. They found that proteasomes, particularly the immunoproteasome, preferentially generated N-terminally rather than C-terminally extended analogs of the SL8 peptide. This study supports the in vivo studies discussed above and directly implicates the proteasomes in generating the C-termini of antigenic peptides. It is important to note that biochemical proof that the C-termini of antigenic peptides are generated in the cytosol based upon analysis of proteolytic intermediates in the cytosol is not yet available. Furthermore, the requirement that the C terminus be made in the cytoplasm does not appear to be absolute. This is illustrated by the discovery of MHC I–bound peptides that were derived from signal sequences (14, 95) and by the presentation of three different peptides from the HIV env protein in TAP deficient T2 cells (96, 97). The location of these peptides

4 Feb 2002

11:41

480

AR

AR152-16.tex

SHASTRI

¥

AR152-16.SGM

SCHWAB

¥

LaTeX2e(2001/05/10)

P1: GKD

SERWOLD

requires liberation of both N- and C-terminal flanking residues, indicating that both N- and C-terminal trimming could occur in the secretory pathway.

Annu. Rev. Immunol. 2002.20:463-493. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

OTHER CYTOPLASMIC PROTEASES Cytoplasmic proteases other than proteasomes may also be involved in generating peptides for MHC I molecules. This evidence comes from mass-spectrometric analysis of peptide generation in proteasome-inhibited cells and from a number of T cell epitopes whose expression increases in the absence of the proteasome (21, 79, 80). Based upon its ability to cleave peptides in vitro and its upregulation in cells surviving partial proteasome inhibition, tripeptidylpeptidase II (TPPII) was suggested to be a peptide supplier (85). In addition, several proteases with aminopeptidase activity have been proposed to take part in the antigen processing pathway. Leucine aminopeptidase was found to generate actual antigenic peptides when potential N-terminally extended precusors were treated in vitro (98). Notably, the expression of leucine aminopeptidase is upregulated in γ -interferon treated cells. Recently two other aminopeptidases were identified by purification from the cytosol (99). Each of these proteases, puromycin sensitive aminopeptidase and bleomycin hydrolase, was able to trim the N-termini of synthetic peptide substrates. However, the role of these proteases in the antigen processing pathway has not yet been established and will require protease-deficient cells or highly specific inhibitors.

ARGUMENTS FAVORING THE GENERATION OF C-TERMINI IN THE CYTOSPLASM Regardless of the mechanism, it appears that the C-terminus of the antigenic peptide must be made in the cytoplasm because proteases in the ER do not seem to be capable of C-terminal trimming. This was evident in an early study by Eisenlohr et al., who found that an antigenic precursor with two extra C-terminal flanking residues could not be presented by MHC I unless the cells also co-expressed a carboxypeptidase in the ER (100). Subsequent analyses with several different antigenic peptides have shown that N-terminal but not C-terminal extensions can be efficiently trimmed from ER-targeted antigenic precursors (101–103). Independently the lack of carboxypeptidase activity in the ER was also inferred from the analysis of the rat cim effect (104). The rat RT1.Aa MHC molecule prefers the arginine residue at the peptide’s C-terminus. RT1.Aa could not be loaded with such peptides in cells that expressed the TAP2B allele because it transports peptides with hydrophobic C-termini, and these could not be trimmed back to arginine in the ER. Finally, direct quantitative analysis of naturally processed peptides showed that the ER is about 300-fold less efficient in trimming C- rather than N-terminal flanking residues from an antigenic peptide (23). Together these studies strongly support the idea that C-termini are made in the cytoplasm because they are inefficiently generated in the ER.

4 Feb 2002

11:41

AR

AR152-16.tex

AR152-16.SGM

LaTeX2e(2001/05/10)

P1: GKD

Annu. Rev. Immunol. 2002.20:463-493. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

LOADING NATURE’S GENE-CHIPS

481

The second attractive feature of generating the C-termini of antigenic peptides in the cytoplasm is that it could explain the intriguing link between TAP specificity and the conserved C-termini of MHC binding motifs. In the mouse and rat cimb strain, for example, TAP prefers to transport peptides with hydrophobic C-termini, mirroring the C-terminal preference of the available MHC I molecules. Whereas the C-terminal bias of TAP is consistent with MHC I–binding motifs, the length preference of TAP is variable. Peptides significantly longer than binding motifs allowed are efficiently transported, suggesting that further trimming, possibly at the N-terminus, might be required in the ER (105, 106). Analysis of TAP transport specificity in vivo is also consistent with this analysis (107). Therefore, a general correspondence between the C-terminus specificity of the proteasome TAP and MHC binding underscores the importance of cytoplasmic events in generating the C-termini of peptides presented by MHC I.

SHUTTLING THE INTERMEDIATES How the cytoplasmic intermediates en route to TAP get there is not known. Peptides may simply diffuse or may be specifically shuttled assisted by chaperones. Several independent studies have shown that cytosolic and ER heat shock proteins hsp70, hsc73, hsp90, and gp96 are associated with antigenic peptides and can elicit donor APC–specific CTL responses (reviewed in 108). These studies have unequivocally shown that these chaperones are associated with antigenic peptides, but the functional role of these chaperones in the antigen processing pathway is not clear. Recently Binder et al. showed that cells presented specific p/MHC I with a high efficiency when peptide-loaded hsp70 or hsp90 were introduced into the cytoplasm (109). Furthermore, cells treated with deoxyspergualin, which binds these heat-shock proteins, disrupted antigen presentation. Similar findings were also reported by T. Torigoe and N. Sato (personal communication). In addition they determined that hsc73 can physically associate with TAP, providing a potential role for peptide-loaded heat-shock proteins in peptide delivery to TAP. This may increase the p/MHC I presentation efficiency by decreasing nonspecific degradation in the cytosol. Because peptide association with the chaperones can serve as an important check-point in the antigen processing pathway and influence the overall efficiency of p/MHC I expression, this is an important area of research and is reviewed by Srivastava in this volume.

THE “START” AT THE END: AMINOPEPTIDASES GENERATE THE N-TERMINUS IN THE ER There is now good evidence indicating that for most peptides, the C-terminus is generated in the cytoplasm. The N-terminal trimming necessary to generate the start of the antigenic peptide appears to be less constrained and may occur in either the cytoplasm or the ER. Several recent studies have established the

4 Feb 2002

11:41

Annu. Rev. Immunol. 2002.20:463-493. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

482

AR

AR152-16.tex

SHASTRI

¥

AR152-16.SGM

SCHWAB

¥

LaTeX2e(2001/05/10)

P1: GKD

SERWOLD

general importance of ER aminopeptidases for the generation of many antigenic peptides. Whereas we refer to these events as occurring in the ER, we emphasize that this location has not been firmly established. The studies most pertinent to this issue have been carried out either in isolated microsomes (as representative of the ER) or in cells lacking TAP, in which antigenic precursors are introduced into the ER via the translocon. It is conceivable that microsomes may be contaminated with other organelles and that antigenic precursors introduced into the ER may have been processed in post-ER compartments and returned into the ER by retrograde transport (110, 111). Therefore, strictly speaking the site of processing in these studies is actually a post-TAP compartment that we refer to here as the ER because MHC I are located and loaded in this compartment, as well as for the sake of brevity. As reviewed above, studies using ER-targeted precursors have shown that Nterminal but not C-terminal flanking residues can be efficiently cleaved from antigenic peptides in the ER. However, because the precursors used in these studies were targeted into the ER necessarily bypassing TAP, it was unclear whether these precursors were trimmed by proteases that were an integral part of the normal antigen processing pathway. Three recent studies have addressed this issue. Paz et al. performed a direct biochemical analysis of the intermediates to the SL8 peptide using their novel HPLC/enzymatic method. They found that N-terminally extended intermediates, and not the minimal peptide, were transported into the ER in living cells. In the presence of Kb, these precursors were trimmed and high levels of the Kb/SL8 complex were generated (74). In a second study, Lavau et al. showed that two antigenic peptides presented by HLA-A2 were poor substrates for TAP in vitro, and when expressed as minimal precursors in the cytoplasm they were not presented (112). Interestingly, extending these precursors to include either one or two N-terminal amino acids greatly enhanced the ability of these peptides to be transported by TAP, and similar extensions of the minimal precursors enabled in vivo presentation. Finally, Serwold et al. treated cells with the aminopeptidase inhibitor leucinethiol, which blocked peptide trimming in the ER, and they found a 30–70% reduction of surface MHC I expression, clearly showing that aminopeptidase trimming in the ER is an essential component of the antigen processing pathway (23). Serwold et al. also systematically defined the specificity of the ER aminopeptidase using a series of ER targeted precursors with N-terminal extensions (23). All N-terminal amino acids, with the important exception of proline, could be efficiently trimmed from the precursors. Instead of being trimmed to the minimal peptide, proline-flanked precursors were trimmed to yield products with the “X-Pro” sequence at their N-terminus (X-P-Xn). Interestingly, the expression of Ld, which prefers X-P-Xn peptides, was severely reduced in cells treated with leucinethiol. This is consistent with the idea that the generation of the X-P-Xn requires aminopeptidase activity and that these peptides are used by MHC I. Aminopeptidase trimming of precursors in the ER to preferentially generate X-P-Xn peptides provided a plausible solution to another paradox in the antigen

4 Feb 2002

11:41

AR

AR152-16.tex

AR152-16.SGM

LaTeX2e(2001/05/10)

P1: GKD

Annu. Rev. Immunol. 2002.20:463-493. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

LOADING NATURE’S GENE-CHIPS

483

processing pathway. TAP translocation assays have consistently shown that peptides containing proline in position 2 are the least preferred substrates for TAP (43, 57). Therefore, it was rather surprising that this “untransportable” sequence should account for approximately 20% of known MHC I-binding motifs (27, 56). Serwold’s observation that aminopeptidase trimming leads to the preferential accumulation of X-P-Xn peptides explained the source of these peptides and also suggested that MHC I molecules that bind to the X-P-Xn motif may have evolved to take advantage of this abundant pool of peptides. An interesting corollary is the possibility that TAP evolved to avoid transporting X-P-Xn peptides specifically because these could not be further tailored. Interestingly, the “X-P” sequence is found not only among peptides loaded onto MHC I, but also among a number of other ER resident proteins in the antigen processing pathway. Tapasin, calreticulin, and β2M, as well as some MHC I alleles share the same X-P residues at their N-termini, suggesting that they too might need protection from a highly active aminopeptidase in the vicinity of the peptide loading complex. Together these studies provide compelling arguments for aminopeptidase trimming in the ER as a key event in the antigen processing pathway. However, the molecular mechanism of ER trimming is obscure. An interesting candidate was recently suggested by the Srivastava group, who discovered that gp96, an ERresident chaperone, possesses aminopeptidase activity (113). This is an intriguing observation because gp96 is well known for its ability to bind antigenic peptides (108). The authors note, however, that the gp96 aminopeptidase activity is extremely weak, a million–fold lower than that of the benchmark aminopeptidase M. Whether low enzymatic activity rules out gp96 as the putative ER aminopeptidase or whether its high abundance compensates for its low activity remains to be determined.

MHC MOLECULES AS TEMPLATES FOR N-TERMINAL TRIMMING IN THE ER The notion that the N-terminal flanking residues of many peptides are trimmed in the ER raises the question of how this trimming is accomplished. In the context of the Rammensee paradox, the MHC I molecules could define the peptide pool in the ER by two different mechanisms (64). The proteolytic intermediates could be processively trimmed at the N-terminus independently of the MHC I. The product of each cleavage cycle could be tested for binding to the MHC I molecules so that once the optimal peptide is made it could be firmly bound and remain protected from further trimming. Alternatively, the extended intermediates could associate with the peptide receptive MHC I molecules, which could then serve as a template for the aminopeptidase until the optimal N-terminus was achieved. Whereas the influence of MHC I on the peptide pool is clear, distinguishing these two models has remained difficult. In one study a nonapeptide (M-SL8) and an octapeptide (SL8) were shown to bind Kb on the cell surface (69). In

4 Feb 2002

11:41

Annu. Rev. Immunol. 2002.20:463-493. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

484

AR

AR152-16.tex

SHASTRI

¥

AR152-16.SGM

SCHWAB

¥

LaTeX2e(2001/05/10)

P1: GKD

SERWOLD

addition, when M-SL8 was expressed as a precursor in cells, it was presented by Db, showing that it was present in the ER. Nevertheless, only SL8, but not MSL8, was recovered from Kb expressing cells, suggesting the possibility that the M-SL8 nonapeptide was first bound to Kb in the ER and was then trimmed to the SL8 peptide. Alternatively, it was possible that a quality control mechanism prevented M-SL8/Kb binding in the ER but not on the cell surface. Stronger evidence favoring a role of MHC I molecules in trimming peptides was noted by Paz et al. (74). In cells without Kb Paz et al. found an N-terminally extended K-SL8 peptide in ER extracts. Remarkably, in cells expressing Kb this precursor decreased dramatically with a concomitant increase in the accumulation of the Kb-bound SL8 peptide. The simplest interpretation of this result is that the N-terminally extended K-SL8 peptide was a precursor for SL8 and that its conversion required the Kb MHC molecule. A prediction of the template model of antigen processing is that MHC I molecules should be bound to extended peptides before the final peptide product is generated. Brouwenstijn in our laboratory recently discovered such complexes in isolated murine microsomes pulsed with N-terminally extended precursors (114). These extended peptide/MHC I complexes were rapidly converted to the optimal peptide/MHC I complexes by an aminopeptidase activity. The discovery of this intermediate together with the Paz and Serwold studies discussed above provide a satisfactory resolution to the Rammensee paradox. They explain why the naturally processed peptides are not found in cells expressing the antigen without the MHC I. The antigen processing mechanism provides only N-terminally extended peptides to the ER that cannot be normally detected owing to the sensitivity limits of the methods. Whether similar extended peptide-bound MHC I complexes exist in vivo is not known. Also, the mechanism of template-driven trimming of extended peptides and the identity of the aminopeptidase have not yet been defined.

A GENERAL MODEL FOR THE P/MHC I ANTIGEN PROCESSING PATHWAY In conclusion, based upon the evidence discussed above, we suggest a model for the MHC I antigen processing pathway. Newly synthesized polypeptides including cryptic translation products serve as the primary source of antigenic peptides. The precursors are cleaved by the proteasome and other proteases to generate a mixture of proteolytic intermediates that have the correct C-terminal end of the antigenic peptide. The intermediates, perhaps assisted by cytoplasmic chaperones, are transported by TAP into the ER. After transport into the ER, the extra Nterminal residues are processively removed by an aminopeptidase, likely using the MHC as a template, until the product fits the antigen-binding groove of the MHC I molecule with high affinity. This final peptide/MHC I complex exits the ER and serves as a potential ligand for CD8+ T cells on the APC surface.

4 Feb 2002

11:41

AR

AR152-16.tex

AR152-16.SGM

LaTeX2e(2001/05/10)

P1: GKD

Annu. Rev. Immunol. 2002.20:463-493. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

LOADING NATURE’S GENE-CHIPS

485

Several features of this model make it attractive for explaining the efficiency of the antigen processing pathway. First, using newly synthesized precursors rather than protein turnover as the primary source of antigenic peptides allows the antigen processing mechanism to tap into the total range of polypeptides synthesized in the cells, regardless of when and where the mature proteins are degraded. This mechanism would also accelerate the recognition of abnormal cells. Second, the stepwise proteolysis of precursors has important implications for the efficiency of peptide generation. Cytoplasmic processing in this view yields a rough-draft of the final peptide with the correct C-terminus. Cytoplasmic processing and transport, by making the C-termini, essentially define which peptides can be presented. Nonetheless, the proteolytic mechanisms do not bear the burden of generating the exact peptides that the MHC I molecules need in the ER. ER trimming of N-terminal flanking residues, under the guidance of MHC I molecules, enables the mechanism to be flexible, and at the same time, maximizes precision. ACKNOWLEDGMENTS We acknowledge our colleagues, past and present, who contributed methods, key observations, and ideas discussed here. We thank Natalie Brouwenstijn for help with the artwork. We are also grateful to colleagues who shared their findings prior to publication. Research in this laboratory is supported by grants from the NIH to NS. Visit the Annual Reviews home page at www.annualreviews.org

LITERATURE CITED 1. Cresswell P, Bangia N, Dick T, Diedrich G. 1999. The nature of the MHC class I peptide loading complex. Immunol. Rev. 172:21–28 2. van Endert PM. 1999. Genes regulating MHC class I processing of antigen. Curr. Opin. Immunol. V11:82–88 3. Grandea AG, Van Kaer L. 2001. Tapasin: an ER chaperone that controls MHC class I assembly with peptide. Trends Immunol. 22:194–99 4. Spiliotis ET, Osorio M, Zuniga MC, Edidin M. 2000. Selective export of MHC class I molecules from the ER after their dissociation from TAP. Immunity 13:841– 51 5. Yewdell JW, Norbury CC, Bennink JR. 1999. Mechanisms of exogenous antigen presentation by MHC class I molecules in vitro and in vivo: implications for gener-

6.

7.

8.

9.

10.

ating CD8+ T cell responses to infectious agents, tumors, transplants, and vaccines. Adv. Immunol. 73:1–77 Watts C, Amigorena S. 2000. Antigen traffic pathways in dendritic cells. Traffic 1:312–17 Larsson M, Fonteneau JF, Bhardwaj N. 2001. Dendritic cells resurrect antigens from dead cells. Trends Immunol. 22:141– 48 Mellman I, Steinman RM. 2001. Dendritic cells: specialized and regulated antigen processing machines. Cell 106:255–58 Rock KL, Rothstein L, Gamble S. 1990. Generation of class I MHC-restricted T-T hybridomas. J. Immunol. 145:804–11 Karttunen J, Shastri N. 1991. Measurement of ligand induced activation in single viable T-cells using the lacZ reporter gene. Proc. Natl. Acad. Sci. USA 88:3972–76

4 Feb 2002

11:41

Annu. Rev. Immunol. 2002.20:463-493. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

486

AR

AR152-16.tex

SHASTRI

¥

AR152-16.SGM

SCHWAB

¥

LaTeX2e(2001/05/10)

P1: GKD

SERWOLD

11. Sanderson S, Shastri N. 1994. LacZ inducible peptide/MHC specific T-hybrids. Int. Immunol. 6:369–76 12. Wallny HJ, Rammensee HG. 1990. Identification of classical minor histocompatibility antigen as cell-derived peptide. Nature 343:275–78 13. R¨otzschke O, Falk K, Wallny HJ, Faath S, Rammensee HG. 1990. Characterization of naturally occurring minor histocompatibility peptides including H-4 and H-Y. Science 249:283–87 14. Henderson RA, Michel H, Sakaguchi K, Shabanowitz J, Appella E, Hunt DF, Engelhard VH. 1992. HLA-A2.1-associated peptides from a mutant cell line: a second pathway of antigen presentation. Science 255:1264–66 15. Henderson RA, Cox AL, Sakaguchi K, Appella E, Shabanowitz J, Hunt DF, Engelhard VH. 1993. Direct identification of an endogenous peptide recognized by multiple HLA-A2.1-specific cytotoxic T cells. Proc. Natl. Acad. Sci. USA 90:10275–79 16. Meadows L, Wang W, den Haan JMM, Blokland E, Reinhardus C, Drijfhout JW, Shabanowitz J, Pierce R, Agulnik AI, Bishop CE, Hunt DF, Goulmy E, Engelhard VH. 1997. The HLAA∗ 0201-restricted H-Y antigen contains a posttranslationally modified cysteine that significantly affects T cell recognition. Immunity 6:273–81 17. Skipper JCA, Hendrickson RC, Gulden PH, Brichard V, Van Pel A, Chen Y, Shabanowitz J, Wolfel T, Slingluff CL, Boon T, Hunt DF, Engelhard VH. 1996. An HLA-A2 restricted tyrosinase antigen on melanoma cells results from post-translational modification and suggests a distinct antigen processing pathway for membrane proteins. J. Exp. Med. 183:527–34 18. Zarling AL, Ficarro SB, White FM, Shabanowitz J, Hunt DF, Engelhard VH. 2000. Phosphorylated peptides are naturally processed and presented by ma-

19.

20.

21.

22.

23.

24.

25.

26.

jor histocompatibility complex class I molecules in vivo. J. Exp. Med. 192:1755– 62 Porgador A, Yewdell JW, Deng Y, Bennink JR, Germain RN. 1997. Localization, quantitation, and in situ detection of specific peptide-MHC class I complexes using a monoclonal antibody. Immunity 6:715–26 Storkus WJ, Zeh HJ III, Salter RD, Lotze MT. 1993. Identification of T-cell epitopes: rapid isolation of class I-presented peptides from viable cells by mild acid elution. J. Immunother. 14:94–103 Luckey CJ, King GM, Marto JA, Venketeswaran S, Maier BF, Crotzer VL, Colella TA, Shabanowitz J, Hunt DF, Engelhard VH. 1998. Proteasomes can either generate or destroy MHC class I epitopes: evidence for nonproteasomal epitope generation in the cytosol. J. Immunol. 161:112–21 Shastri N, Serwold T, Paz P. 1998. Reading within the lines: naturally processed peptides displayed by MHC class I molecules. Curr. Opin. Immunol. 10:137– 44 Serwold T, Gaw S, Shastri N. 2001. ER aminopeptidases generate a unique pool of peptides for MHC class I molecules. Nat. Immunol. 2:644–51 Levitskaya J, Coram M, Levitsky V, Imreh S, Steigerwald-Mullen PM, Klein G, Kurilla MG, Masucci MG. 1995. Inhibition of antigen processing by the internal repeat region of the Epstein-Barr virus nuclear antigen-1. Nature 375:685–88 Blake N, Lee S, Redchenko I, Thomas W, Steven N, Leese A, SteigerwaldMullen P, Kurilla MG, Frappier L, Rickinson A. 1997. Human CD8+ T cell responses to EBV EBNA1: HLA class I presentation of the (Gly-Ala)-containing protein requires exogenous processing. Immunity 7:791–802 Gilbert MJ, Riddell SR, Plachter B, Greenberg PD. 1996. Cytomegalovirus selectively blocks antigen processing and

4 Feb 2002

11:41

AR

AR152-16.tex

AR152-16.SGM

LaTeX2e(2001/05/10)

P1: GKD

LOADING NATURE’S GENE-CHIPS

27.

Annu. Rev. Immunol. 2002.20:463-493. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

28.

29.

30.

31.

32.

33.

34.

35.

presentation of its immediate-early gene product. Nature 383:720–22 Rammensee HG, Bachmann J, Stevanovic S. 1997. MHC Ligands and Peptide Motifs. Austin, TX: Landes Bioscience. 462 pp. Moore MW, Carbone FR, Bevan MJ. 1988. Introduction of soluble protein into the class I pathway of antigen processing and presentation. Cell 54:777–85 Pamer EG, Harty JT, Bevan MJ. 1991. Precise prediction of a dominant class I MHC-restricted epitope of Listeria monocytogenes. Nature 353:852–55 Loveland B, Wang C-R, Yonekawa H, Hermel E, Lindahl KF. 1990. Maternally transmitted histocompatibility antigen of mice: a hydrophobic peptide of a mitochondrially encoded protein. Cell 60:971– 80 Bhuyan PK, Young LL, Fischer Lindhal K, Butcher GW. 1997. Identification of the rat maternally transmitted minor histocompatibility antigen. J. Immunol. 158:3753–60 Bacik I, Snyder HL, Anton LC, Russ G, Chen W, Bennink JR, Urge L, Otvos L, Dudkowska B, Eisenlohr L, Yewdell JW. 1997. Introduction of a glycosylation site into a secreted protein provides evidence for an alternative antigen processing pathway: transport of precursors of major histocompatibility complex class I-restricted peptides from the endoplasmic reticulum to the cytosol. J. Exp. Med. 186:479– 87 Varshavsky A. 1996. The N-end rule: functions, mysteries, uses. Proc. Natl. Acad. Sci. USA 93:12142–49 Grant EP, Michalek MT, Goldberg AL, Rock KL. 1995. Rate of antigen degradation by the ubiquitin-proteasome pathway influences MHC class I presentation. J. Immunol. 155:3750–58 Sijts AJAM, Pilip I, Pamer EG. 1997. The Listeria monocytogenes-secreted p60 protein is an N-end rule substrate in the cytosol of infected cells: implications for

36.

37.

38.

39.

40.

41.

42.

43.

487

major histocompatibility complex class antigen processing of bacterial proteins. J. Biol. Chem. 272:19261–68 Townsend A, Bastin J, Gould K, Brownlee G, Andrew M, Coupar B, Boyle D, Chan S, Smith G. 1988. Defective presentation to class I-restricted cytotoxic T lymphocytes in vaccinia-infected cells is overcome by enhanced degradation of antigen. J. Exp. Med. 168:1211–24 Gueguen M, Long EO. 1996. Presentation of a cytosolic antigen by major histocompatibility complex class II molecules requires a long-lived form of the antigen. Proc. Natl. Acad. Sci. USA 93:14692– 97 Goth S, Nguyen V, Shastri N. 1996. Generation of naturally processed peptide/ MHC class I complexes is independent of the stability of endogenously synthesized precursors. J. Immunol. 157:1894–904 Reits EAJ, Vos JC, Gromme M, Neefjes J. 2000. The major substrates for TAP in vivo are derived from newly synthesized proteins. Nature 404:774–78 Schubert U, Anton LC, Gibbs J, Norbury CC, Yewdell JW, Bennink JR. 2000. Rapid degradation of a large fraction of newly synthesized proteins by proteasomes. Nature 404:770–74 Yewdell JW, Anton LC, Bennink JR. 1996. Defective ribosomal products (DRiPs): a major source of antigenic peptides for MHC class I molecules? J. Immunol. 157:1823–26 Ossendorp F, Eggers M, Neisig A, Ruppert T, Groettrup M, Sijts A, Mengede E, Kloetzel PM, Neefjes J, Koszinowski U, Melief C. 1996. A single residue exchange within a viral CTL epitope alters proteasome-mediated degradation resulting in lack of antigen presentation. Immunity 5:115–24 Neisig A, Roelse J, Sijts AJAM, Ossendorp F, Feltkamp MCW, Kast WM, Melief CJM, Neefjes JJ. 1995. Major differences in transporter associated with antigen presentation (TAP)-dependent

4 Feb 2002

11:41

488

44.

Annu. Rev. Immunol. 2002.20:463-493. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

45.

46.

47.

48.

49.

50.

51.

52.

AR

AR152-16.tex

SHASTRI

¥

AR152-16.SGM

SCHWAB

¥

LaTeX2e(2001/05/10)

P1: GKD

SERWOLD

translocation of MHC class I-presentable peptides and the effect of flanking sequences. J. Immunol. 154:1273–79 Falk K, Rotzschke O, Stevanovic S, Jung G, Rammensee H-G. 1991. Allelespecific motifs revealed by sequencing of self-peptides eluted from MHC molecules. Nature 351:290–96 Boon T, Van Pel A. 1989. T cellrecognized antigenic peptides derived from the cellular genome are not protein degradation products but can be generated directly by transcription and translation of short subgenic regions. A hypothesis. Immunogenetics 29:75–79 Iborra FJ, Jackson DA, Cook PR. 2001. Coupled transcription and translation within nuclei of mammalian cells. Science V293:1139–42 Bullock TNJ, Eisenlohr LC. 1996. Ribosomal scanning past the primary initiation codon as a mechanism for expression of CTL epitopes encoded in alternative reading frames. J. Exp. Med. 184:1319– 29 Shastri N, Nguyen V, Gonzalez F. 1995. Major histocompatibility class I molecules can present cryptic translation products to T-cells. J. Biol. Chem. 270:1088–91 Malarkannan S, Horng T, Shih P, Schwab S, Shastri N. 1999. Presentation of out-offrame peptide/MHC class I complexes by a novel translation initiation mechanism. Immunity 10:681–90 Madden DR, Gorga JC, Strominger JL, Wiley DC. 1991. The structure of HLAB27 reveals nonamer self-peptides bound in an extended conformation. Nature 353: 321–25 Madden DR. 1995. The three-dimensional structure of peptide-MHC complexes. Annu. Rev. Immunol. 13:587–622 Fremont DH, Matsumura M, Stura EA, Peterson PA, Wilson IA. 1992. Crystal structures of two viral peptides in complex with murine MHC class I H-2Kb. Science 257:919–27

53. Collins EJ, Garboczi DN, Wiley DC. 1994. Three-dimensional structure of a peptide extending from one end of a class I MHC binding site. Nature 371:626–29 54. Speir JA, Stevens J, Joly E, Butcher GW, Wilson IA. 2001. Two different, highly exposed, bulged structures for an unusually long peptide bound to rat MHC class I RT1-Aa. Immunity 14:81–92 55. Hennecke J, Wiley DC. 2001. T cell receptor-MHC interactions up close. Cell 104:1–4 56. Barber LD, Gillece-Castro B, Percival L, Li X, Clayberger C, Parham P. 1995. Overlap in the repertoires of peptides bound in vivo by a group of related class I HLA-B allotypes. Curr. Biol. 5:179– 90 57. Van Endert PM, Riganelli D, Greco G, Fleischhauer K, Sidney J, Sette A, Bach JF. 1995. The peptide-binding motif for the human transporter associated with antigen processing. J. Exp. Med. 182: 1883–95 58. Sykulev Y, Joo M, Vturina I, Tsomides TJ, Eisen HN. 1996. Evidence that a single peptide-MHC complex on a target cell can elicit a cytolytic T cell response. Immunity 4:565–71 59. Hunt DF, Henderson RA, Shabanowitz J, Sakaguchi K, Michel H, Sevilir N, Cox AL, Appella E, Engelhard VH. 1992. Characterization of peptides bound to the class I MHC molecule HLA-A2.1 by mass spectrometry. Science 255:1261–63 60. Cox AL, Skipper J, Chen Y, Henderson RA, Darrow TL, Shabanowitz J, Engelhard VH, Hunt DF, Slingluff CL Jr. 1994. Identification of a peptide recognized by five melanoma-specific human cytoxic T cell lines. Science 264:716–19 61. Malarkannan S, Gonzalez F, Nguyen V, Adair G, Shastri N. 1996. Alloreactive CD8+ T cells can recognize unusual, rare, and unique processed peptide/MHC complexes. J. Immunol. 157:4464–73 62. Mendoza L, Paz P, Zuberi AR, Christianson G, Roopenian DC, Shastri N.

4 Feb 2002

11:41

AR

AR152-16.tex

AR152-16.SGM

LaTeX2e(2001/05/10)

P1: GKD

LOADING NATURE’S GENE-CHIPS

63.

Annu. Rev. Immunol. 2002.20:463-493. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

64.

65.

66.

67.

68.

69.

70.

1997. Minors held by majors. The H13 minor histocompatibility locus defined as a peptide/MHC class I complex. Immunity 7:461–72 Van Bleek GM, Nathenson SG. 1990. Isolation of an endogenously processed immunodominant viral peptide from the class I H-2Kb molecule. Nature 348:213– 16 Falk K, R¨otzschke O, Rammensee HG. 1990. Cellular peptide composition governed by major histocompatibility complex class I molecules. Nature 348:248– 51 R¨otzschke O, Falk K, Deres K, Schild H, Norda M, Metzger J, Jung G, Rammensee HG. 1990. Isolation and analysis of naturally processed viral peptides as recognized by cytotoxic T cells. Nature 348:252–54 Nieland TJF, Agnes MC, Tan A, Monnee-Van Muijen M, Koning F, Kruisbeek AM, Van Bleek GM. 1996. Isolation of an immunodominant viral peptide that is endogenously bound to the stress protein GP96-GRP94. Proc. Natl. Acad. Sci. USA 93:6135–39 Del Val M, Schlicht H-J, Ruppert T, Reddehase MJ, Koszinowski UH. 1991. Efficient processing of an antigenic sequence for presentation by MHC class I molecules depends upon its neighboring residues in the protein. Cell 66:1145– 53 Griem P, Wallny H-J, Falk K, Rotschke O, Arnold B, Schonrich G, Hammerling G, Rammensee H-G. 1991. Uneven tissue distribution of minor histocompatibility proteins versus peptides is caused by MHC expression. Cell 65:633– 40 Malarkannan S, Goth S, Buchholz DR, Shastri N. 1995. The role of MHC class I molecules in the generation of endogenous peptide/MHC complexes. J. Immunol. 154:585–98 Buchholz D, Scott P, Shastri N. 1995. Presentation without processing of en-

71.

72.

73.

74.

75.

76.

77.

78.

79.

489

dogenous precursors in the MHC class I presentation pathway. J. Biol. Chem. 270:6515–22 Anton LC, Yewdell JW, Bennink JR. 1997. MHC class I-associated peptides produced from endogenous gene products with vastly different efficiencies. J. Immunol. 158:2535–42 Elliott T, Townsend A, Cerundolo V. 1990. Antigen presentation: naturally processed peptides. Nature 348:195–97 Dick TP, Ruppert T, Groettrup M, Kloetzel PM, Kuehn L, Koszinowski UH, Stevanovic S, Schild H, Rammensee HG. 1996. Coordinated dual cleavages induced by the proteasome regulator PA28 lead to dominant MHC ligands. Cell 86:253–62 Paz P, Brouwenstijn N, Perry R, Shastri N. 1999. Discrete proteolytic intermediates in the MHC class I antigen processing pathway and MHC I-dependent peptide trimming in the ER. Immunity 11:241–51 Serwold T, Shastri N. 1999. Specific proteolytic cleavages limit the diversity of the pool of peptides available to MHC class I molecules in living cells. J. Immunol. 162:4712–19 Brown GM, Driscoll J, Monaco JJ. 1991. Structural and serological similarity of MHC-linked LMP and proteasome (multicatalytic proteinase) complexes. Nature 353:355–57 Rock KL, Gramm C, Rothstein L, Clark K, Stein R, Dick L, Hwang D, Goldberg AL. 1994. Inhibitors of the proteasome block the degradation of most cell proteins and the generation of peptides presented on MHC class I molecules. Cell 78:761– 71 Rock KL, Goldberg AL. 1999. Degradation of cell proteins and the generation of MHC class I-presented peptides. Annu. Rev. Immunol. 17:739–79 Vinitsky A, Anton LC, Snyder HL, Orlowski M, Bennink JR, Yewdell JW. 1997. The generation of MHC class I-associated peptides is only partially

4 Feb 2002

11:41

490

Annu. Rev. Immunol. 2002.20:463-493. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

80.

81.

82.

83.

84.

85.

86.

87.

88.

AR

AR152-16.tex

SHASTRI

¥

AR152-16.SGM

SCHWAB

¥

LaTeX2e(2001/05/10)

P1: GKD

SERWOLD

inhibited by proteasome inhibitors: involvement of nonproteasomal cytosolic proteases in antigen processing? J. Immunol. 159:554–64 Luckey CJ, Marto JA, Partridge M, Hall E, White FM, Lippolis JD, Shabanowitz J, Hunt DF, Engelhard VH. 2001. Differences in the expression of human class I MHC alleles and their associated peptides in the presence of proteasome inhibitors. J. Immunol. 167:1212–21 Niedermann G, Geier E, Lucchiari-Hartz M, Hitziger N, Ramsperger A, Eichmann K. 1999. The specificity of proteasomes: impact on MHC class I processing and presentation of antigens. Immunol. Rev. 172:29–48 Kloetzel P-M. 2001. Antigen processing by the proteasome. Nat. Rev. Mol. Cell Biol. 2:179–87 Monaco JJ, Nandi D. 1995. The genetics of proteasomes and antigen processing. Annu. Rev. Genet. 29:729–54 Glas R, Bogyo M, McMaster JS, Gaczynska M, Ploegh HL. 1998. A proteolytic system that compensates for loss of proteasome function. Nature 392:618– 22 Geier E, Pfeifer G, Wilm M, LucchiariHartz M, Baumeister W, Eichmann K, Niedermann G. 1999. A giant protease with potential to substitute for some functions of the proteasome. Science 283:978– 81 Van Kaer L, Ashton-Rickardt PG, Eichelberger M, Gaczynska M, Nagashima K, Rock KL, Goldberg AL, Doherty P, Tonegawa S. 1994. Altered peptidase and viral specific T cell response in LMP2 mutant mice. Immunity 1:533– 41 Fehling HJ, Swat W, Laplace C, Kuhn R, Rajewsky K, Muller U, von Boehmer H. 1994. MHC class I expression in mice lacking the proteasome subunit LMP-7. Science 265:1234–37 Preckel T, Fung-Leung W-P, Cai Z, Vitiello A, Salter-Cid L, Winqvist O, Wolfe

89.

90.

91.

92.

93.

94.

95.

96.

TG, Von Herrath M, Angulo A, Ghazal P, Lee J-D, Fourie AM, Wu Y, Pang J, Ngo K, Peterson PA, Fruh K, Yang Y. 1999. Impaired immunoproteasome assembly and immune responses in PA28-/mice. Science 286:2162–65 Van Kaer L, Ashton-Rickardt PG, Ploegh HL, Tonegawa S. 1992. TAP1 mutant mice are deficient in antigen presentation, surface class I molecules, and CD4-8+ T cells. Cell 71:1205–14 Grandea AG, Golovina TN, Hamilton SE, Sriram V, Spies T, Brutkiewicz RR, Harty JT, Eisenlohr LC, Van Kaer L. 2000. Impaired assembly yet normal trafficking of MHC class I molecules in tapasin mutant mice. Immunity 13:213– 22 Garbi N, Tan P, Diehl AD, Chambers BJ, Ljunggren H-G, Momburg F, Haemmerling GJ. 2000. Impaired immune responses and altered peptide repertoire in tapasin-deficient mice. Nat. Immunol. 1:234–38 Craiu A, Akopian T, Goldberg A, Rock KL. 1997. Two distinct proteolytic processes in the generation of a major histocompatibility complex class Ipresented peptide. Proc. Natl. Acad. Sci. USA 94:10850–55 Mo XY, Cascio P, Lemerise K, Goldberg AL, Rock K. 1999. Distinct proteolytic processes generate the C and N termini of MHC class I-binding peptides. J. Immunol. 163:5851–59 Cascio P, Hilton C, Kisselev AF, Rock KL, Goldberg AL. 2001. 26S proteasomes and immunoproteasomes produce mainly N-extended versions of an antigenic peptide. EMBO J. V20:2357– 66 Wei ML, Cresswell P. 1992. HLA-A2 molecules in an antigen-processing mutant cell contain signal sequence-derived peptides. Nature 356:443–46 Hammond SA, Bollinger RC, Tobery TW, Siliciano RF. 1993. Transporterindependent processing of Hiv-1 envelope

4 Feb 2002

11:41

AR

AR152-16.tex

AR152-16.SGM

LaTeX2e(2001/05/10)

P1: GKD

LOADING NATURE’S GENE-CHIPS

Annu. Rev. Immunol. 2002.20:463-493. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

97.

98.

99.

100.

101.

102.

103.

104.

protein for recognition by CD8-positive T cells. Nature 364:158–61 Hammond SA, Johnson RP, Kalams SA, Walker BD, Takiguchi M, Safrit JT, Koup RA, Siliciano RF. 1995. An epitope-selective, transporter associated with antigen presentation (TAP)-1/2-independent pathway and a more general TAP-1/2-dependent antigen-processing pathway allow recognition of the HIV-1 envelope glycoprotein by CD8+ CTL. J. Immunol. 154:6140–56 Beninga J, Rock KL, Goldberg AL. 1998. Interferon-gamma can stimulate post-proteasomal trimming of the N terminus of an antigenic peptide by inducing leucine aminopeptidase. J. Biol. Chem. 273:18734–42 Stoltze L, Schirle M, Schwarz G, Schroeter C, Thompson MW, Hersh LB, Kalbacher H, Stevanovic S, Rammensee H-G, Schild H. 2000. Two new proteases in the MHC class I processing pathway. Nat. Immunol. 1:413–18 Eisenlohr LC, Bacik I, Bennink JR, Bernstein K, Yewdell JW. 1992. Expression of a membrane protease enhances presentation of endogenous antigens to MHC class I-restricted T lymphocytes. Cell 71:963–72 Snyder HL, Yewdell JW, Bennink JR. 1994. Trimming of antigenic peptides in an early secretory compartment. J. Exp. Med. 180:2389–94 Snyder HL, Bacik I, Bennink JR, Kearns G, Behrens TW, Bachi T, Orlowski M, Yewdell JW. 1997. Two novel routes of transporter associated with antigen processing (TAP)-independent major histocompatibility complex class I antigen processing. J. Exp. Med. 186:1087– 98 Elliott T, Willis A, Cerudolo V, Townsend A. 1995. Processing of major histocompatibility class I-restricted antigens in the endoplasmic reticulum. J. Exp. Med. 181:1481–91 Powis SJ, Young LL, Joly E, Barker

105.

106.

107.

108.

109.

110.

111.

112.

491

PJ, Richardson L, Brandt RP, Melief CJ, Howard JC, Butcher GW. 1996. The Rat cim Effect: TAP allele-dependent changes in a class I MHC anchor motif and evidence against C-terminal trimming of peptides in the ER. Immunity 4:159– 65 Heemels MT, Ploegh H. 1995. Generation, translocation, and presentation of MHC class I-restricted peptides. Annu. Rev. Biochem. 64:463–91 Androlowicz MJ, Cresswell P. 1996. How selective is the transporter associated with antigen processing? Immunity 5:1–5 Yellen-Shaw A, Lughlin CE, Metrione RM, Eisenlohr LC. 1997. Murine transporter associated with antigen presentation (TAP) preferences influence class Irestricted T cell responses. J. Exp. Med. 186:1655–62 Srivastava PK, Menoret A, Basu S, Binder RJ, McQuade KL. 1998. Heat shock proteins come of age: primitive functions acquire new roles in an adaptive world. Immunity 8:657–65 Binder RJ, Blachere NE, Srivastava PK. 2001. Heat shock protein-chaperoned peptides but not free peptides introduced into the cytosol are presented efficiently by major histocompatibility complex I molecules. J. Biol. Chem. V276:17163– 71 Townsend A, Ohlen C, Bastin J, Ljunggren H-G, Foster L, Karre K. 1989. Association of class I major histocompatibility heavy and light chains induced by viral peptides. Nature 340:443–48 Day PM, Yewdell JW, Porgador A, Germain RN, Bennink JR. 1997. Direct delivery of exogenous MHC class I moleculebinding oligopeptides to the endoplasmic reticulum of viable cells. Proc. Natl. Acad. Sci. USA 94:8064–69 Lauvau G, Kakimi K, Niedermann G, Ostankovitch M, Yotnda P, Firat H, Chisari FV, van Endert PM. 1999. Human transporters associated with antigen processing (TAPs) select epitope precursor

4 Feb 2002

11:41

492

113.

Annu. Rev. Immunol. 2002.20:463-493. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

114.

115.

116.

117.

118.

119.

120.

AR

AR152-16.tex

SHASTRI

¥

AR152-16.SGM

SCHWAB

¥

LaTeX2e(2001/05/10)

P1: GKD

SERWOLD

peptides for processing in the endoplasmic reticulum and presentation to T cells. J. Exp. Med. 190:1227–40 Menoret A, Li Z, Niswonger ML, Altmeyer A, Srivastava PK. 2001. An endoplasmic reticulum protein implicated in chaperoning peptides to major histocompatibility of class I is an aminopeptidase. J. Biol. Chem. 276:33313–8 Brouwenstijn N, Serwold T, Shastri N. 2001. MHC class I molecules can direct proteolytic cleavage of antigenic precursors in the endoplasmic reticulum. Immunity 15:95–104 Gooding LR, O’Connell KA, Geib RW, Pipas JM. 1988. Cytotoxic T cell recognition of the SV40 tumor antigen: a note of caution. In Processing and Presentation of Antigens, ed. B Pernis, S Silverstein, H Vogel, pp. 87–95. New York: Academic Boon T, Van Pel A, De Plaen E, Chomez P, Lurquin C, Szikora J-P, Sibille C, Mariame B, Van Den Eynde B, Lethe B, Brichard V. 1989. Genes coding for Tcell-defined tum transplantation antigens: point mutations, antigenic peptides, and subgenic expression. Cold Spring Harbor Symp. Quant. Biol. 54:587–96 Fetten JV, Roy N, Gilboa E. 1991. A frameshift mutation at the NH2 terminus of the nucleoprotein gene does not affect generation of cytotoxic T lymphocyte epitopes. J. Immunol. 147:2697–705 Hahn YS, Braciale VL, Braciale TJ. 1991. Presentation of viral antigen to class I major histocompatibility complexrestricted cytotoxic T lymphocyte. Recognition of an immunodominant influenza hemagglutinin site by cytotoxic T lymphocyte is independent of the position of the site in the hemagglutinin translation product. J. Exp. Med. 174:733–36 Shastri N, Gonzalez F. 1993. Endogenous generation and presentation of the OVA peptide/Kb complex to T-cells. J. Immunol. 150:2724–36 Uenaka A, Ono T, Akisawa T, Wada H, Yasuda T, Nakayama E. 1994. Identifica-

121.

122.

123.

124.

125.

126.

127.

tion of a unique antigen peptide pRL1 on BALB-c RL1 leukemia recognized by cytotoxic T lymphocytes and its relation to the Akt oncogene. J. Exp. Med. 180:1599– 607 Coulie PG, Lehmann F, Lethe B, Herman J, Lurquin C, Andrawiss M, Boon T. 1995. A mutated intron sequence codes for an antigenic peptide recognized by cytolytic T lymphocytes on a human melanoma. Proc. Natl. Acad. Sci. USA 92:7976–80 Malarkannan S, Afkarian M, Shastri N. 1995. A rare cryptic translation product is presented by Kb major histocompatibility complex class I molecule to alloreactive T cells. J. Exp. Med. 182:1739–50 Elliott T, Bodmer H, Townsend A. 1996. Recognition of out-of-frame major histocompatibility complex class Irestricted epitopes in vivo. Eur. J. Immunol. 26:1175–79 Guilloux Y, Lucas S, Brichard VG, Van Pel A, Viret C, De Plaen E, Brasseur F, Lethe B, Jotereau F, Boon T. 1996. A peptide recognized by human cytolytic T lymphocytes on HLA-A2 melanomas is encoded by an intron sequence of the Nacetylglucosaminyltransferase V gene. J. Exp. Med. 183:1173–83 Wang R-F, Parkhurst MR, Kawakami Y, Robbins PF, Rosenberg SA. 1996. Utilization of an alternative open reading frame of a normal gene in generating a novel human cancer antigen. J. Exp. Med. 183:1131–40 Bullock TNJ, Patterson AE, Franlin LL, Notidis E, Eisenlohr LC. 1997. Initiation codon scanthrough versus termination codon readthrough demonstrates strong potential for major histocompatibility complex class I-restricted cryptic epitope expression. J. Exp. Med. 186:1051–58 Robbins PF, El-Gamil M, Li YF, Fitzgerald EB, Kawakami Y, Rosenberg SA. 1997. The intronic region of an incompletely spliced gp100 gene transcript

4 Feb 2002

11:41

AR

AR152-16.tex

AR152-16.SGM

LaTeX2e(2001/05/10)

P1: GKD

LOADING NATURE’S GENE-CHIPS

Annu. Rev. Immunol. 2002.20:463-493. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

128.

129.

130.

131.

encodes an epitope recognized by melanoma-reactive tumor-infiltrating lymphocytes. J. Immunol. 159:303–8 Lupetti R, Pisarra P, Verrecchia A, Farina C, Nicolini G, Anichini A, Bordignon C, Sensi M, Parmiani G, Traversari C. 1998. Translation of a retained intron in tyrosinase-related protein (TRP) 2 mRNA generates a new cytotoxic T lymphocyte (CTL)-defined and shared human melanoma antigen not expressed in normal cells of the melanocytic lineage. J. Exp. Med. 188:1005–16 Mayrand SM, Schwarz DA, Green WR. 1998. An alternative translational reading frame encodes an immunodominant retroviral CTL determinant expressed by an immunodeficiency-causing retrovirus. J. Immunol. 160:39–50 Shichijo S, Nakao M, Imai Y, Takasu H, Kawamoto M, Niiya F, Yang D, Toh Y, Yamma H, Itoh K. 1998. A gene encoding antigenic peptides of human squamous cell carcinoma recognized by cytotoxic T lymphocytes. J. Exp. Med. 187:277–88 Wang R-F, Johnston SL, Zeng G, Topalian SL, Schwartzentruber DJ, Rosenberg S. 1998. A breast and melanoma-shared tumor antigen: T cell responses to antigenic peptides translated from different open reading frames. J. Immunol. 161: 3596–606

493

132. Aarnoudse CA, van den Doel PB, Heemskerk B, Schrier PI. 1999. Interleukin2-induced, melanoma-specific T cells recognize CAMEL, an unexpected translation product of LAGE-1. Int. J. Cancer 82:442–48 133. Dolstra H, Fredrix H, Maas F, Coulie PG, Brasseur F, Mensink E, Adema GJ, de Witte TM, Figdor CG, Wielvan VD, Kemenade E. 1999. A human minor histocompatibility antigen specific for B cell acute lymphoblastic leukemia. J. Exp. Med. 189:301–8 134. Ronsin C, Chung-Scott V, Poullion I, Aknouche N, Gaudin C, Triebel F. 1999. A non-AUG-defined alternative open reading frame of the intestinal carboxyl esterase mRNA generates an epitope recognized by renal cell carcinomareactive tumor-infiltrating lymphocytes in situ. J. Immunol. 163:483–90 135. Probst-Kepper M, Stroobant V, Kridel R, Gaugler B, Landry C, Brasseur F, Cosyns JP, Weynand B, Boon T, Van Den Eynde BJ. 2001. An alternative open reading frame of the human macrophage colony-stimulating factor gene is independently translated and codes for an antigenic peptide of 14 amino acids recognized by tumor-infiltrating CD8 T lymphocytes. J. Exp. Med. 193:1189– 98

P1: FRK

February 1, 2002

12:47

Annual Reviews

AR152-FM

Annual Review of Immunology Volume 20, 2002

Annu. Rev. Immunol. 2002.20:463-493. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

CONTENTS FRONTISPIECE, Charles A. Janeway, Jr. A TRIP THROUGH MY LIFE WITH AN IMMUNOLOGICAL THEME, Charles A. Janeway, Jr.

xiv 1

THE B7 FAMILY OF LIGANDS AND ITS RECEPTORS: NEW PATHWAYS FOR COSTIMULATION AND INHIBITION OF IMMUNE RESPONSES, Beatriz M. Carreno and Mary Collins

29

MAP KINASES IN THE IMMUNE RESPONSE, Chen Dong, Roger J. Davis, and Richard A. Flavell

PROSPECTS FOR VACCINE PROTECTION AGAINST HIV-1 INFECTION AND AIDS, Norman L. Letvin, Dan H. Barouch, and David C. Montefiori T CELL RESPONSE IN EXPERIMENTAL AUTOIMMUNE ENCEPHALOMYELITIS (EAE): ROLE OF SELF AND CROSS-REACTIVE ANTIGENS IN SHAPING, TUNING, AND REGULATING THE AUTOPATHOGENIC T CELL REPERTOIRE, Vijay K. Kuchroo, Ana C. Anderson, Hanspeter Waldner, Markus Munder, Estelle Bettelli, and Lindsay B. Nicholson

55 73

101

NEUROENDOCRINE REGULATION OF IMMUNITY, Jeanette I. Webster, Leonardo Tonelli, and Esther M. Sternberg

125

MOLECULAR MECHANISM OF CLASS SWITCH RECOMBINATION: LINKAGE WITH SOMATIC HYPERMUTATION, Tasuku Honjo, Kazuo Kinoshita, and Masamichi Muramatsu

165

INNATE IMMUNE RECOGNITION, Charles A. Janeway, Jr. and Ruslan Medzhitov

KIR: DIVERSE, RAPIDLY EVOLVING RECEPTORS OF INNATE AND ADAPTIVE IMMUNITY, Carlos Vilches and Peter Parham ORIGINS AND FUNCTIONS OF B-1 CELLS WITH NOTES ON THE ROLE OF CD5, Robert Berland and Henry H. Wortis E PROTEIN FUNCTION IN LYMPHOCYTE DEVELOPMENT, Melanie W. Quong, William J. Romanow, and Cornelis Murre

197 217 253 301

LYMPHOCYTE-MEDIATED CYTOTOXICITY, John H. Russell and Timothy J. Ley

SIGNAL TRANSDUCTION MEDIATED BY THE T CELL ANTIGEN x

323

P1: FRK

Annu. Rev. Immunol. 2002.20:463-493. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

February 1, 2002

12:47

Annual Reviews

AR152-FM

CONTENTS

xi

RECEPTOR: THE ROLE OF ADAPTER PROTEINS, Lawrence E. Samelson INTERACTION OF HEAT SHOCK PROTEINS WITH PEPTIDES AND ANTIGEN PRESENTING CELLS: CHAPERONING OF THE INNATE AND ADAPTIVE IMMUNE RESPONSES, Pramod Srivastava CHROMATIN STRUCTURE AND GENE REGULATION IN THE IMMUNE SYSTEM, Stephen T. Smale and Amanda G. Fisher PRODUCING NATURE’S GENE-CHIPS: THE GENERATION OF PEPTIDES FOR DISPLAY BY MHC CLASS I MOLECULES, Nilabh Shastri, Susan

371

Schwab, and Thomas Serwold

395 427

463

THE IMMUNOLOGY OF MUCOSAL MODELS OF INFLAMMATION, Warren Strober, Ivan J. Fuss, and Richard S. Blumberg T CELL MEMORY, Jonathan Sprent and Charles D. Surh

GENETIC DISSECTION OF IMMUNITY TO MYCOBACTERIA: THE HUMAN MODEL, Jean-Laurent Casanova and Laurent Abel ANTIGEN PRESENTATION AND T CELL STIMULATION BY DENDRITIC CELLS, Pierre Guermonprez, Jenny Valladeau, Laurence Zitvogel, Clotilde Th´ery, and Sebastian Amigorena

495 551 581

621

NEGATIVE REGULATION OF IMMUNORECEPTOR SIGNALING, Andr´e Veillette, Sylvain Latour, and Dominique Davidson

669

CPG MOTIFS IN BACTERIAL DNA AND THEIR IMMUNE EFFECTS, Arthur M. Krieg

PROTEIN KINASE Cθ

709 IN

T CELL ACTIVATION, Noah Isakov and Amnon

Altman

RANK-L AND RANK: T CELLS, BONE LOSS, AND MAMMALIAN EVOLUTION, Lars E. Theill, William J. Boyle, and Josef M. Penninger PHAGOCYTOSIS OF MICROBES: COMPLEXITY IN ACTION, David M. Underhill and Adrian Ozinsky

761 795 825

STRUCTURE AND FUNCTION OF NATURAL KILLER CELL RECEPTORS: MULTIPLE MOLECULAR SOLUTIONS TO SELF, NONSELF DISCRIMINATION, Kannan Natarajan, Nazzareno Dimasi, Jian Wang, Roy A. Mariuzza, and David H. Margulies

853

INDEXES Subject Index Cumulative Index of Contributing Authors, Volumes 1–20 Cumulative Index of Chapter Titles, Volumes 1–20

ERRATA An online log of corrections to Annual Review of Immunology chapters (if any, 1997 to the present) may be found at http://immunol.annualreviews.org/

887 915 925

8 Feb 2002

10:2

AR

AR152-17.tex

AR152-17.SGM

LaTeX2e(2001/05/10) P1: GKD 10.1146/annurev.immunol.20.100301.064816

Annu. Rev. Immunol. 2002. 20:495–549 DOI: 10.1146/annurev.immunol.20.100301.064816

THE IMMUNOLOGY OF MUCOSAL MODELS OF INFLAMMATION1 Warren Strober,1 Ivan J. Fuss,1 and Richard S. Blumberg2 Annu. Rev. Immunol. 2002.20:495-549. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

1

Mucosal Immunity Section, Laboratory of Clinical Investigation, NIAID, NIH, Bethesda, Maryland 20892-1890; e-mail: [email protected] 2 Division of Gastroenterology, Brigham and Women’s Hospital, Harvard Medical School, Boston, Massachusetts 02115

Key Words Crohn’s disease, ulcerative colitis, tolerance, cytokines, Th1/Th2 ■ Abstract In recent years the status of the inflammatory bowel diseases (IBDs) as canonical autoimmune diseases has risen steadily with the recognition that these diseases are, at their crux, abnormalities in mucosal responses to normally harmless antigens in the mucosal microflora and therefore responses to antigens that by their proximity and persistence are equivalent to self-antigens. This new paradigm is in no small measure traceable to the advent of multiple models of mucosal inflammation whose very existence is indicative of the fact that many types of immune imbalance can lead to loss of tolerance for mucosal antigens and thus inflammation centered in the gastrointestinal tract. We analyze the immunology of the IBDs through the lens of the murine models, first by drawing attention to their common features and then by considering individual models at a level of detail necessary to reveal their individual capacities to provide insight into IBD pathogenesis. What emerges is that murine models of mucosal inflammation have given us a road map that allows us to begin to define the immunology of the IBDs in all its complexity and to find unexpected ways to treat these diseases.

INTRODUCTION The study of animal models of mucosal inflammation as a means to probe the pathogenesis of inflammatory bowel disease (IBD) extends back almost a half century [for reviews of the older literature see Strober (1) and Kim & Berstad (2)], and it is fair to say that this kind of study embodied the first serious attempt to determine the immunologic basis of this category of disease. One class of early models is that devised by Kirsner and his colleagues in the early 1960s in which the mucosal immune system was manipulated in some way to cause a 1 The US Government has the right to retain a nonexclusive, royalty-free license in and to any copyright covering this paper.

495

4 Feb 2002

11:44

Annu. Rev. Immunol. 2002.20:495-549. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

496

AR

AR152-17.tex

STROBER

¥

FUSS

AR152-17.SGM

¥

LaTeX2e(2001/05/10)

P1: GKD

BLUMBERG

mucosal (colonic) inflammation (3). Perhaps the most interesting of these models consisted of colonic inflammation that was induced in rabbits via the “Auer procedure,” wherein rabbits are first immunized with an antigen (such as OVA), then subjected to disruption of the colonic epithelial barrier with formalin, and finally are re-administered the original antigen by a local mucosal or system route (4). This procedure led to a colonic inflammation not unlike that in ulcerative colitis but was transient even when the procedure was repeated in the same animal. A more sustained inflammation, however, was obtained in the late 1970s by Mee et al., who modified the Auer procedure by sensitizing animals (rabbits) to an Escherichia coli–associated antigen (5). Similarly, in studies performed some 10 years earlier, now almost forgotten, Halpern et al. showed that immunization of rats with live or dead E. coli (in Freund’s adjuvant) led to chronic colitis even without introduction of a colonic irritant per rectum; in addition, feeding of E. coli prevented the development of colitis (6). These studies, together with early studies of dinitrochlorobenzene-induced colitis reported about the same time as the studies of Mee et al. and coworkers (5, 7, 8), clearly indicated that an initial immunologic assault of varying cause on the gastrointestinal tract can lead to more sustained inflammation as a result of a break in normal “tolerance” to antigens in the mucosal microflora. Another class of early models of mucosal inflammation were those produced by physical agents and included colitides produced by exposure to acetic acid, phorbol ester, F-met-leu-phe, and various sulfated polysaccharides such as carageenan, amylopectin sulfate, and dextran sulfate sodium (DSS) (9–20). One common feature of these agents appears to be their capacity to disrupt the epithelial cell barrier and therefore to promote increased cellular exposure to normal mucosal microflora. Evidence for this comes from studies of DSS-induced colitis in which it has been shown that DSS alters mucosal barrier function prior to the onset of colitis (19). In addition, colitis caused by exposure to F-met-leu-phe has also been shown to be associated with changes in barrier function, in this case mediated by neutrophils (14). One possible or even probable consequence of this change in barrier function is that mucosal phagocytes become subject to activation by substances in the mucosal flora and this, in turn, leads to antigen nonspecific release of pro-inflammatory cytokines (e.g., TNF-α) and inflammation. This scenario is supported by the observation that both DSS colitis and carageenan colitis can be effectively treated with antibiotics (20, 21). Disruption of barrier function(s) as a mechanism in physical agent–induced colitis fits with a second common feature of colitides caused by physical agents, namely their relative independence from lymphocyte-mediated responses. Thus, in DSS-induced colitis, it is evident that mice lacking T cells, B cells, and NK cells can still develop colitis in response to DSS (22). This being said, in the presence of an intact immune system containing these cellular elements, dextran sulfate leads to activation of lymphocytes and the induction of Th1 and/or Th2 responses. This leads to the conclusion that in physical agent–induced colitis a T cell–mediated inflammation can be superimposed on macrophage-induced inflammation.

4 Feb 2002

11:44

AR

AR152-17.tex

AR152-17.SGM

LaTeX2e(2001/05/10)

P1: GKD

Annu. Rev. Immunol. 2002.20:495-549. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

INSIGHTS INTO IBD

497

This brief survey of historically important models of mucosal inflammation leaves little doubt that on close reading they provide data that presaged many of the findings obtained with a set of newer models that had been extensively characterized over the past decade. Thus, as alluded to above, these models revealed the critical role of the mucosal microflora in the pathogenesis of mucosal inflammation and the related role of barrier function as a bulwark against extensive stimulation of the mucosal immune system by the microflora. In addition, they provide the first insights into the often overlooked role of antigen nonspecific factors in mucosal inflammation and even provide an early hint of the role of active tolerance in preventing such inflammation [the feeding studies of Halpern et al. (6)]. These considerations, of course, are by no means meant to belittle the new knowledge of mucosal inflammation (and by extension IBD) that have come from studies of newer models. Not only have the latter provided for the first time a detailed framework for the understanding of the various proinflammatory and antiinflammatory mechanisms at work in this type of inflammation, but they have also provided us with invaluable clues as to how the latter can be effectively treated in humans. In discussing these newer models of mucosal inflammation, we first survey their common features to derive basic principles of mucosal inflammation that are applicable to this area of study as a whole. We then discuss major individual models in depth, emphasizing the particular insights derivable from each model and how each model helps establish the basic principles and mechanisms of mucosal inflammation. This, the main body of the review, is subdivided into sections on Th1 models, Th2 models, and barrier function models.

BASIC (GENERAL) FEATURES OF MODELS OF MUCOSAL INFLAMMATION As is evident from the detailed review of individual models of mucosal inflammation that follows, certain recurrent principles emerge that relate to all models to a greater or lesser extent. These principles together define the basic immunology of both models of mucosal inflammation and of human IBD, and thus it is useful to discuss them first in an outline form that can later be fleshed out in the discussions of individual models to follow.

Final Common Pathways of Mucosal Inflammation As becomes amply evident below, models of mucosal inflammation reflect a remarkably wide variety of causes. Nevertheless, the resulting inflammation that develops is almost always channeled into a final common pathway of inflammation, mediated by either an excessive Th1 T-cell response associated with excessive IL-12/IFN-γ /TNF-α secretion or an excessive Th2 T-cell response associated with increased IL-4/IL-5 secretion (reviewed in 23). The great majority of models are in fact Th1 models, but why this is the case is far from clear (see Table 1). One factor may relate to the influence of strain on disease because the given model may

4 Feb 2002

11:44

498

AR

AR152-17.tex

STROBER

¥

FUSS

AR152-17.SGM

¥

LaTeX2e(2001/05/10)

P1: GKD

BLUMBERG

TABLE 1 Models of mucosal inflammation classified by nature of T cell-mediated inflammation Th1 Models

Th2 Models

TNBS colitis (SJL/J mice)

TCR-α chain deficiency

SCID-transfer colitis

TNBS colitis in BALB/c mice**

TCR Tg mice with lymphopenia a

IL-10 deficiency colitis

Oxazalone colitis WASP deficiency

Annu. Rev. Immunol. 2002.20:495-549. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

IL-10 signaling defects (CRF2-4 deficiency) Tgε26 mice TNF1ARE mice (TNF-α overproduction) C3H/HeJBir mice Gi2α-deficient mice Samp1/Yit mice T-bet Tg mice STAT4 Tg mice TGF-β RII dominant-negative Tg mice HLA-B27 Tg rats Mdr1a-deficient mice DSS colitis IL-7 Tg mice a

Abbreviations: SCID, severe combined immunodeficiency; TCR, T cell receptor; CRF2-4, cyto receptor family 2-4; TNF, tumor necrosis factor; STAT-4, signal transduction and activators of transcription-4; TGF, transforming growth factor; DSS, dextran sulfate sodium; WASP, Wiskott-Aldrich syndrome protein. **Mixed response but initially Th1, later Th2. Abbreviations: TNBS, trinitrobenzene sulfonic acid; SCID, xx; TCR, T cell receptor; CRF2-4, xx; TNF, xx; STAT4, xx; TGF, xx; HLA-B27, xx; DSS, xx; WASP, xx.

manifest a Th1 character in the SJL/J strain mouse but may manifest a Th2 character (or mixed Th1/Th2 character) in the BALB/c mice. A more likely explanation, however, relates to the fact that in most if not all models the inflammation is driven by antigens in the normal mucosal microflora, which in effect means that it will be influenced by mitogens [e.g., lipopolysaccharides (LPS), CpGs] and superantigens associated with these organisms that tend to induce IL-12 production and thus Th1 responses. This is nicely illustrated by trinitrobenzene sulfonic acid (TNBS)– colitis in SJL/J mice that manifest increased LPS-driven IL-12 responses, which are thought to play a key role in the Th1 response elicited by TNBS administration (see discussion below; G. Bouma & W. Strober, unpublished observations). A final possibility is that many of the models are due to a failure to regulate mucosal responses that are essentially normal responses to antigens in the mucosal milieu. As we see below, such regulation most likely involves the secretion of

4 Feb 2002

11:44

AR

AR152-17.tex

AR152-17.SGM

LaTeX2e(2001/05/10)

P1: GKD

Annu. Rev. Immunol. 2002.20:495-549. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

INSIGHTS INTO IBD

499

suppressor cytokines such as TGF-β (or IL-10), which is more closely related to the regulation of Th1 responses than Th2 responses (24, 25). In fact, there is considerable evidence that Th1 responses suppress both the expansion of TGF-β– secreting cells and TGF-β signaling (26, 27) and, contrariwise, TGF-β interferes with IL-12 signaling (28–30). Thus, Th1 and TGF-β responses have a reciprocal relationship to one another and appear to be mutually exclusive. On the other hand, there is some evidence that Th2 and TGF-β responses can co-exist and that it requires higher levels of TGF-β to suppress a Th2 response than a Th1 response (25, 31). These considerations suggest that Th1-mediated mucosal inflammations are more sensitive to defects of regulation (mediated by TGF-β) and thus, defects in regulation will more frequently lead to Th1-mediated inflammation than Th2mediated inflammation. The bias of the experimental models toward Th1-mediated inflammation raises the question as to when and how Th2-mediated inflammation ever occurs. One factor is again the nature of the antigen driving the inflammation or, alternatively, the specificity of the T cell receptor (TCR) on the reactive T cells. In this regard, certain antigens are “Th2-type antigens,” perhaps because the nature of the antigen dictates the type of antigen presenting cell that induces T cell differentiation in Th2 T cells. Evidence for this is inherent in the fact that one haptenating agent, TNBS, elicits a Th1 response in SJL/J mice, whereas another, oxazalone, elicits a Th2 response (31, 32). In addition, a Th2-oriented response may result in colitis associated with TCR-α chain deficiency because in this situation the T cells utilize a TCR (a ββ TCR) that may have the ability to recognize and expand in response to antigens only under conditions that allow Th2 responses (33, 34). One thing to keep in mind, however, is that the Th1-mediated inflammation may switch to a Th2 inflammation under some circumstances. This is seen in IL-10–deficient mice, perhaps because in the absence of IL-10, cells in which IL-4 signaling leads to GATA-3 suppression of IL-12 signaling gradually accumulate, and ultimately a Th2 T cell dominates the inflammation (35; A.D. Levine, personal communication). Whether a Th1 or a Th2 response is responsible for the mucosal inflammation has considerable impact on the nature of the inflammation because, as we see below, Th1 responses are marked by transmural cellular infiltration that in some cases is associated with granulomata (i.e., TNF1ARE model and SAMP1/Yit model) (36–38), and whereas epithelial cell layer changes are clearly present, they are not a dominant feature. A similar histopathologic picture is obtained in Crohn’s disease, and thus it is fair to say that, in general, Th1 models are related to this human disease (23). This presumed association between Th1 models and Crohn’s disease is also strengthened by the fact that Crohn’s disease is in fact a Th1-mediated inflammation (39–42). Th2-mediated inflammations are, on the contrary, marked by more superficial cellular infiltrates associated with a greater disruption of the epithelial layer and in some cases greater polymorphonuclear infiltration. This situation is more akin to ulcerative colitis, but this correlation is inexact because ulcerative colitis has not been clearly shown to be a Th2-mediated inflammation. Thus, whereas some authors have found high IL-5 levels in ulcerative colitis, IL-4

4 Feb 2002

11:44

500

AR

AR152-17.tex

STROBER

¥

FUSS

AR152-17.SGM

¥

LaTeX2e(2001/05/10)

P1: GKD

BLUMBERG

levels are quite normal. Clearly, if a Th2 inflammation is present in ulcerative colitis, it is a highly atypical Th2 inflammation (39, 40).

Cellular Elements Involved in Mucosal Inflammation Antigen presenting cells (APCs) in mucosal tissues are probably key cells in the induction of both mucosal effector and regulatory cell responses. Hence, it is likely (but not yet proven) that defects of T cell responses arise either from defects in APC function or APC–T cell interactions. Alternatively, it is possible that APCs are the target of regulatory cells, a possibility proposed by Malmstrom et al. in relation to OX40-positive APCs present in the mesenteric nodes of mice with SCID-transfer colitis (43). Macrophages, a type of APC, are activated in mucosal inflammation and function mainly as effector cells. However, these cells may also be involved in regulatory interactions. This possibility is realized in mice with myeloid cell–specific STAT3 deficiency that have macrophages that cannot produce several STAT3dependent cytokines, such as the important regulatory cytokine, IL-10 (44). Thus, in vitro macrophages in this model of inflammation exhibit heightened effector activity characterized by increased LPS-induced production of IL-12, TNF-α, IL-6, and IL-1β; thus, in vivo these macrophages lead to LPS-induced mucosal inflammation.

Annu. Rev. Immunol. 2002.20:495-549. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

ANTIGEN PRESENTING CELLS

T CELLS T cells play multiple roles in experimental mucosal inflammation both as effector cells and regulatory cells. The former are mainly CD4+ T cells because these cells make up the main cell populations that infiltrate mucosal tissues in all models so far studied and because in instances in which they are deleted in vivo, inflammation is ameliorated (45). CD8+ T cells are also present in tissues but do not appear to play a decisive pathologic role because in the few instances in which they were deleted in vivo no major effect on inflammation was obtained (46). This does not, however, rule out a supportive pathogenic role because increased cytotoxic T cell function has been observed in some of the models (47). Evidence has recently appeared that indicates that loss of epithelial cells in ulcerative colitis can be attributed to a T cell or NK cell–mediated cytotoxic event (48). The above information on the role of cytotoxicity in models would suggest, however, that even if cytotoxic elimination of epithelial cells occurs in ulcerative colitis, such cytotoxicity is not likely to be a major component of the overall immunopathologic process. γ δ T cells, i.e., T cells confined to the intra-epithelial compartment, do not play an important role as effector cells in any form of colitis, except perhaps in TCR-α chain–deficient mice, in which they are present in increased numbers (49). In recent studies it has been shown that whereas γ δ T cells cannot in themselves induce colitis (in lymphopenic TCR-α chain–deficient mice), injection of anti-TCRδ antibody into TCR-α chain–deficient mice prevented development of colitis (50). On this basis, γ δ T cells in this context appear to play an accessory

4 Feb 2002

11:44

AR

AR152-17.tex

AR152-17.SGM

LaTeX2e(2001/05/10)

P1: GKD

Annu. Rev. Immunol. 2002.20:495-549. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

INSIGHTS INTO IBD

501

role in the inflammation. γ δ T cells also do not have a major regulatory role in mucosal inflammation, as γ chain–deficient mice do not develop inflammation (51). Here too, however, one can find an exception, in that mice lacking γ δ T cells are reported to develop more severe TNBS-colitis (52). Finally, one very definite and in this case positive role of γ δ T cells in mucosal inflammation is their role in the healing of mucosal inflammation. This is shown by the fact that intra-epithelial γ δ T cells produce factors, notably keratinocyte growth factor, that may facilitate restoration of epithelial cell barrier integrity in DSS-colitis (53). Whereas, as indicated above, CD4+ T cells can function in the various models as either Th1 or Th2 effector cells, they can also function as regulatory cells. With regard to the latter, several different types of cells have been described, but it is very possible that these are in reality one cell that appears in different disguises. One type of regulatory cell is a TGF-β–secreting T cell (a so-called Th3 cell), which is the cell induced by antigen feeding during the development of oral tolerance. The mucosal cytokine milieu necessary for the induction of this cell is not well understood, although it is known that Th2 conditions favor induction and Th1 conditions inhibit induction (reviewed in 24, 26, 27). IL-10 has been seriously considered as a possible inductive cytokine for this cell, but in recent studies of oral tolerance induction as well as in in vitro studies of Th3 T cell development from naive cells, IL-10 has no direct inductive effect on the development of Th3 T cells and may enhance TGF-β production only through its capacity to down-regulate Th1 responses. However, in the same in vitro studies TGF-β itself had a positive autocrine effect on its own secretion (54). A second type of regulatory cell is an IL-10 secreting cell (a Tr1 cell), which may also secrete small amounts of TGF-β (55). This cell has poor proliferative capacities and in initial studies was induced by sequential antigenic restimulation in the presence of IL-10. More recently, however, it has been shown that both IL-10 and IFN-α are necessary for its induction (56). Yet another regulatory cell is the CD25+ T cell, which is a thymus-derived cell that inhibits effector T cells via cell-cell contact rather than secretion of an inhibitory cytokine (57, 58). Recently, however, Nakamura et al. have reported data that show that most CD25+ T cells bear surface TGF-β in the form of a latent TGF-β protein (TGF-β associated with latency-associated protein) and secrete TGF-β and IL-10 when activated in the presence of IL-2 and/or strong costimulatory signals (59). These authors suggest that under minimal stimulation conditions that might occur prior to inflammation, these cells inhibit via cellular contact and activation of the surface TGF-β at the cell-cell interface. In contrast, they suggest that under maximal stimulation conditions that occur in the presence of inflammation, CD25+ T cells inhibit via secretion of TGF-β and IL-10. Thus, the CD25+ T cells have qualities of both Th3 and Tr1 regulatory cells. A final type of suppressor cell is the NK cell or NK-T cell. The former has been shown to suppress inflammation in the SCID-transfer model of colitis (60), whereas the latter has been shown to suppress inflammation in DSS colitis (61). The NK-T cell preferentially recognizes glycolipid antigens presented via an atypical

4 Feb 2002

11:44

502

AR

AR152-17.tex

STROBER

¥

FUSS

AR152-17.SGM

¥

LaTeX2e(2001/05/10)

P1: GKD

BLUMBERG

Annu. Rev. Immunol. 2002.20:495-549. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

MHC class I molecule (CD1d), which is ordinarily expressed on dendritic cells, B cells, and epithelial cells; thus, this cell may be activated via antigens presented by both epithelial cells and more conventional APCs (62). The mechanism by which NK or NK-T cells suppress mucosal inflammation or other forms of inflammation is poorly understood, as their possible cellular targets of suppression are presently unknown. B CELLS Whereas autoantibodies are found in some models of colitis as well as in human IBD, it does not appear that B cells play a role either in induction of mucosal inflammation or its maintenance. In fact, in the one instance that B cells have been actively studied, in the Th2-mediated inflammation in TCR-α chain deficiency, they appear to play a protective role rather than a pathologic role (see below) (63). Whether such protection also occurs in human IBD is not known.

Epithelial cells form a barrier against exposure to mucosal microflora and other mucosal antigens and thus play a key role in the downregulation of mucosal immune response. As is evident from the discussion of individual models of mucosal inflammation below, in several models alterations in this barrier are the primary cause of colitis (64, 65), whereas in several other models a change in barrier function is a contributory (secondary) factor (66). Epithelial cells also function as sensors of the bacterial microenvironment and release chemokines in a programmed fashion when in contact with pathogens. Such chemokine release has the effect of drawing leukocytes into peri-epithelial sites, which then set up the first line of defense against invading organisms. Whether such chemokine release also plays a role in the initiation of mucosal inflammation is not yet clear. That it may is suggested by a recent study of Mdr1a-deficient mice, i.e., mice whose epithelial cell cannot expel proteins from within the cell including those proteins derived from infectious pathogens (65). As described below, such mice develop colitis that is likely to be due to prolonged secretion of chemokines and cytokines rather than a break in the epithelial cell barrier per se.

EPITHELIAL CELLS

Broad Categories of Mucosal Models of Inflammation As discussed in several previous reviews (24, 67–69), mucosal immune responses are fine-tuned by opposing immune mechanisms that on the one hand lead to effector cell responses addressing host defense at mucosal surfaces and on the other to tolerogenic responses preventing inflammatory reactions to the myriad of antigens in the mucosal environment. It is now apparent that the tolerogenic response has two major components: (a) processes by which mucosal antigens (in the form of unadjuvanted proteins) bring about “classical” tolerance via induction of T cell anergy or deletion either in the mucosal tissues per se or upon gaining entrance to the circulation and the central lymphoid tissues; and (b) processes by which mucosal antigen induces regulatory T cells, which secrete antigen-nonspecific suppressor

4 Feb 2002

11:44

AR

AR152-17.tex

AR152-17.SGM

LaTeX2e(2001/05/10)

P1: GKD

Annu. Rev. Immunol. 2002.20:495-549. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

INSIGHTS INTO IBD

503

cytokines such as IL-10 and TGF-β. These two tolerogenic processes operate in tandem and are probably both necessary to maintain mucosal homeostasis. Thus, whereas induction of anergy/deletion can greatly reduce the number of T cells that can respond to a mucosal antigen, it is probably not able to eliminate all such T cells, and the latter become memory cells potentially able to evoke inflammatory responses. These latter cells are held in check, however, by a cadre of regulatory T cells that respond to the same stimulating antigen. We must bear in mind that the main antigenic pool to which mucosal homeostasis must apply is the antigenic pool associated with the mucosal microflora, which by their persistence and proximity are formally equivalent to self antigens. Given the inevitable inefficiency of a deletion mechanism in relation to so large and mutable an antigenic pool and given the fact discussed above that a full scale effector cell response tends to obliterate a regulatory response (at least for a period of time), the burden (or the challenge) of the mucosal tolerogenic mechanism may fall disproportionately on regulatory cell function. This view is amply supported in the discussions of the individual models below in which the origin of the inflammation can be repeatedly traced to an inadequate regulatory response rather than to a hyperactive or excessive effector response to antigens in the mucosal microflora. The mechanisms governing the development of mucosal tolerance (also called oral tolerance) are not yet completely understood. One important mechanism probably relates to the special nature of the mucosal dendritic cell population, which may have an increased propensity to present antigen in a way that induces either anergic/deletional tolerance or suppressor cell tolerance. Recent work showing that subsets of mucosal dendritic cells have a substantially different cytokine secretion profiles than spleen dendritic cells, i.e., produce more IL-10, supports this concept (70, 71). Nevertheless, much remains to be learned about the origin of these cells and how they shape mucosal responses. In particular, it is not known how these cells are influenced in their development by the adjacent epithelium and how, in turn, these cells induce either the de novo development of regulatory cells or the expansion of a preexisting population of regulatory cells. In any case, the above considerations allow us to classify models of mucosal inflammation into two broad categories (Figure 1): “type 1 models,” wherein the defect lies with the effector mechanisms of the mucosal response and “type 2 models,” wherein the effector cell response is normal, but the regulatory cell response is impaired. One example of a type 1 model is the colitis seen in mice bearing a STAT4 transgene (72). These mice have an increased propensity to mount a Th1 T cell response because of excessive responsiveness to IL-12 signaling; thus, when T cells from these mice are exposed to autologous bacterial antigen in vitro and then transferred to a SCID recipient, they induce colitis in the recipient, whereas naive T cells from normal mice do not. A second example is TNBS colitis in SJL/J mice, wherein it is thought that the colitis is preceded by and is dependent on a genetically determined IL-12 hyperresponsiveness ignited by a disturbance of epithelial barrier function by ethanol followed by exposure of the mucosal APCs

Annu. Rev. Immunol. 2002.20:495-549. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

4 Feb 2002 11:44

504

AR AR152-17.tex

STROBER ¥

FUSS

AR152-17.SGM

¥

BLUMBERG

LaTeX2e(2001/05/10) P1: GKD

4 Feb 2002

11:44

AR

AR152-17.tex

AR152-17.SGM

LaTeX2e(2001/05/10)

P1: GKD

INSIGHTS INTO IBD

505

Annu. Rev. Immunol. 2002.20:495-549. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

TABLE 2 Models of mucosal inflammationa Type I models

Type II models

TNFARE mice

SCID-transfer colitis

TNBS colitis

IL-10 deficiency and IL-10 signaling defect colitis

?C3H/HeJBir mice

?C3H/HeJBir mice

Gi2α-deficient mice

IL-2-deficient mice

STAT4 Tg mice

TGF-β RII dominant-negative mice

N-cadherin dominant-negative mice

Tgε26 mice

IL-7 Tg mice DSS colitis Mice with NF-κB defects a

Unidentified: SAMP1/Yit mice; HLA-B27 Tg rats, mice with Wiskott-Aldrich syndrome protein deficiency.

Abbreviations: TNF, xx; TNBS, trinitrobenzene sulfonic acid; STAT4, xx; DSS, dextran sulfate sodium; SCID, xx; TGF, xx. For other abbreviations, see Table 1.

to antigens in the mucosal microflora (32). This response then conditions the mice to respond to TNBS with a massive Th1 response that rapidly inhibits a normal counter-regulatory response. Type II models, i.e., models that result from an inadequate regulatory response, are exemplified by the SCID-transfer model wherein transfer of naive CD45Rhi T cells leads to colitis, whereas transfer of both naive and memory (CD45RBlow) T cells does not (73). In this model, as described more fully below, the memory cell population contains regulatory cells so that transfer of only naive cells leads to an inadequate regulatory response and colitis. A second type II model that results from inadequate regulation is that seen in mice bearing a dominant-negative TGF-β RII chain (under a CD4+ promotor) that abrogates TGF-β signaling (74, 75). Here, regulatory cells are present, but they cannot function adequately because their intended targets are “blind” to their signals. In the following detailed review of various models, we characterize their basic mechanisms as a type I model (faulty effector cell function) or as a type II model (inadequate regulatory cell function), and in Table 2 we have categorized most of the models on this basis. ←−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−− − Figure 1 (upper panel) The normal mucosal immune system displays a balanced effector T cell response (Th1 or Th2) and regulatory T cell response (Th3 or Tr1). (lower panels) The abnormal mucosal immune system displays an unbalanced response consisting of either excessive effector cell response (type I models) or inadequate regulatory cell response (type II models).

4 Feb 2002

11:44

506

AR

AR152-17.tex

STROBER

¥

FUSS

AR152-17.SGM

¥

LaTeX2e(2001/05/10)

P1: GKD

BLUMBERG

Annu. Rev. Immunol. 2002.20:495-549. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

The Role of Mucosal Microflora in the Induction of Mucosal Inflammation Regardless of whether the experimental mucosal inflammation is type I or type II in character as defined above, the driving force of the inflammation is the nonpathogenic commensal organism resident on the mucosal surface, the mucosal microflora. This is supported by the data in Table 3, which shows that, with perhaps one or two exceptions, mice developing disease in a specific pathogen-free or conventional environment do not do so in a germ-free environment, and in most instances disease is ameliorated when the mice are treated with antibiotics that rid the mucosa of certain classes of organisms (reviewed in 21, 76–85). Exceptions to this consistent pattern are informative. The first exception is the IL-2–deficient mouse, which develops severe and aggressive gastrititis, duodenitis, and colitis under conventional conditions but only nonfatal, mild, focal, and nonproliferative gastrointestinal inflammation under germ-free conditions (80, 81, 86). In addition, these mice develop peri-portal hepatic inflammation, anemia, and generalized lymphoid hyperplasia, which is not ameliorated by the presence of a germ-free state. Thus, this exception to the rule can be explained if we assume that autoimmune inflammation against nonmucosal self-antigens is a component of IL-2 deficiency disease. The second exception is the induced colitis known as dextran sulfate colitis or DSS colitis. This model of colitis can also be observed under germ-free conditions (at least in some studies), although it is ameliorated by antibiotic treatment (21, 83, 84). This can be explained by the fact that this

TABLE 3 Colitis in models of mucosal inflammation in germ-free vs. specific pathogen-free (SPF) or conventional conditions SPF

Germ-free

Antibiotic treatment

SCID-transfer colitis

+

0

c

IL-2 deficiency colitis

++

Mild, focal

IL-10 deficiency colitis

++

0

a

?

Tgε26

+

0

?

TCR-α chain colitis

+

0

? ?

SAMP-1/Yit mice

+

0

DSS colitis

+

0/+b

Carageenan colitis

+

0

Indomethacin colitis

+

0

a

See references 80, 81, 86.

b c

See references 21, 83, 84.

Decreased colitis.

Abbreviations: SCID, xx; TCR, T cell receptor; SAMP, xx; DSS, dextran sulfate sodium.

4 Feb 2002

11:44

AR

AR152-17.tex

AR152-17.SGM

LaTeX2e(2001/05/10)

P1: GKD

Annu. Rev. Immunol. 2002.20:495-549. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

INSIGHTS INTO IBD

507

is a colitis caused primarily by direct activation of macrophages by a physical agent (DSS), and T cell responses appear to be superimposed phenomena that can aggravate but are not essential to the inflammation. Thus, this exception to the rule is due to the fact that the colitis is at least in part driven by nonimmune factors. Additional and more direct evidence that mucosal microflora drive mucosal inflammation in models of mucosal inflammation comes from studies showing that mouse lamina propria T cells are usually unresponsive to their own microbial flora (with respect to either proliferation or cytokine production) but are responsive to the microflora of other individuals even if the other individual is a mouse of the same strain (87, 88). Thus, quite remarkably, oral tolerance to “self flora” appears to be every bit as specific as tolerance to “self antigens.” A related observation is that mice with TNBS colitis lose their nonresponsiveness to their own flora and regain it when the colitis resolves (88). This suggests that the colitis is at least in part driven by the antigen in the microflora, either as a result of cross-reactivity to TNBS or because with the onset of colitis, tolerance to many microflora antigens is lost. Those latter possibilities are supported by the finding that systemic immunization of IL-2–deficient mice with TNP-KLH or other TNP-substituted proteins produces rapid onset of colitis that is identical to the spontaneous colitis occurring in these mice (89). It is important to note that such loss of tolerance to self flora is also a feature of human inflammatory bowel disease (IBD), a fact suggesting that the human disease is also due to loss of tolerance to self microflora (87, 88, 90). A third kind of evidence supporting the fact that antigens in the normal microbial flora drive mucosal inflammation comes from an extensive series of studies of the spontaneous colitis occurring in an LPS-nonresponsive C3H/HeJ mouse substrain (called C3H/HeJBir mice), which is discussed more fully below. Suffice it to say here that CD4+ T cells stimulated in vitro by lysates of resident bacteria can transfer disease to naive disease-free recipients (91, 92). Similarly, studies of mice bearing a STAT4 transgene show that in vitro exposure of T cells from mice with an increased propensity to undergo Th1 T cell differentiation to autologous microfloral antigens induces in these T cells the capacity to cause a Th1 colitis in SCID recipients (72). Together, these studies provide a direct demonstration that T cells specific for mucosal microflora act as effector cells in models of mucosal inflammation. It should be noted, however, that whereas effector cells inducing colitis can be stimulated by antigens in the mucosal microflora, regulatory cells can also be so stimulated. This is shown in additional studies of C3H/HeJBir mice in which it was found that cell lines producing IL-10 could also be derived from these mice, which upon co-transfer with effector cells prevented development of colitis (reviewed in 93). From these studies and other studies below it is evident that the mucosal microflora can also induce regulatory cells, and it is really the loss of balance between induction of effector and regulatory cells that defines when disease occurs. The fact that the mucosal microflora is the major driving force in experimental inflammatory disease should not be taken to imply that all bacterial antigens take part in the disease process or even that the same antigens are necessarily implicated

4 Feb 2002

11:44

Annu. Rev. Immunol. 2002.20:495-549. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

508

AR

AR152-17.tex

STROBER

¥

FUSS

AR152-17.SGM

¥

LaTeX2e(2001/05/10)

P1: GKD

BLUMBERG

in a number of different models of inflammation. This is evident from the studies of the aforementioned C3H/HeJBir model showing that relatively few antigens within the large antigenic pool of the mucosal microflora are actually found to stimulate B cells or T cells during the course of the disease (92). One must hasten to add, however, that while the number of stimulating antigens was low relative to the total number of antigens present, it was still considerable (see below). A similar situation is observed in the SCID-transfer and TCR-α chain–deficient mouse models (34, 94). In the latter, T cells with aberrant T cell receptors (TCRs) exhibit restricted T cell clonality in conjunction with a common (public) motif in the CDR3 region of the TCR (33, 34). However, in this case the aberrant TCR probably dictates a limited ability to recognize the full set of antigens and thus, may exaggerate the narrowness of the antigenic repertoire recognized by the colitic mice. Similar considerations apply to human patients with IBD who also exhibit restricted T cell clonality and evidence of public motifs among the cells present in lesional tissues (95). Thus, in some patients with IBD there is evidence that the restricted T cell clonality reflecting the presence of a limited group of related stimulatory antigens may be involved in disease pathogenesis. However, in the majority of patients the data are more consistent with a broader T cell response that is characterized by the presence of private motifs that vary from individual to individual. The restricted yet variable nature of the antigens of the mucosal microflora capable of evoking mucosal inflammation does not conflict with the fact that monoassociation of HLA-B27 transgenic rats or mice with IL-10 deficiency and TCR-α chain deficiency with Bacteroides vulgatus can lead to colitis (76, 77, 96, 97). First, B. vulgatus is likely to be one among many organisms that can induce disease. Second, the effects of B. vulgatus mono-association may relate to the its ability to synergize with other Enterobacteriaceae in causing infection or to augment internalization of selected strains of bacteria (98). Similarly, Helicobacter hepaticus infection causes disease in IL-10–deficient mice under some animal room conditions but not others (99, 100). In addition, in one study microflora that included H. hepaticus caused colitis in Rag-2–deficient mice but not Rag-2–deficient mice also deficient in IL-7 or Rag-2–deficient mice treated with IL-10 (101). Because overproduction of IL-7 by epithelial cells is a cause of colitis in another model (102), these results suggest that the ability of an organism to cause colitis may depend on its ability to directly stimulate a particular cytokine pattern in the mouse host and thereby cause undue activation of certain cell populations, such as macrophage populations. Studies involving antibiotic treatment of murine models of inflammation with antibiotic also attest to the fact that many bacterial species are capable of promoting inflammation. Thus, whereas either ciprofloxcin or a combination of neomycin and metronizadole could prevent colitis in IL-10–deficient mice, only the combined antibiotic regimen was successful as a treatment of the colitis (85). Similarly, combinations of vancomycin plus imipenem were necessary to treat disease in IL-10–deficient mice, DSS-treated mice, and HLA-B27 transgenic rats, and other

4 Feb 2002

11:44

AR

AR152-17.tex

AR152-17.SGM

LaTeX2e(2001/05/10)

P1: GKD

Annu. Rev. Immunol. 2002.20:495-549. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

INSIGHTS INTO IBD

509

combinations of broad spectrum antibiotics were necessary to treat mice with TNBS colitis (21, 76, 77, 103). In line with the above discussion of effector and regulatory cells stimulated by mucosal microflora, certain organisms appear to be particularly involved in the induction of mucosal inflammation, presumably owing to their capacity to stimulate effector cells, and other organisms may have a special capacity to quell inflammation via an enhanced capacity to stimulate regulatory cells. Evidence in favor of this concept is that introduction of Lactobacillus species into the mucosal environment of IL-10–deficient mice prevents the development of colitis of the mice under specific pathogen-free conditions (104). Additional evidence comes from the observation that whereas colitis in mice with Mdr1a deficiency is worsened by infection with Helicobacter bili, it is ameliorated by infection with H. hepaticus (105). Finally, spontaneous colitis in the SAMP1/Yit mouse is more severe in a pathogen-free environment than in a conventional environment (37). This finding is relevant to human IBD because it is possible that the increase in the incidence of IBD observed in developed countries may be due to the fact that exposure to organisms that could ameliorate potential inflammation is decreased in these countries. This view is consistent with recent studies by Dalwadi et al. showing that a superantigen (called I2) derived from Pseudomonus species is associated with Crohn’s disease lesions and induces the regulatory cytokine IL-10 in vitro (106). Overall, these considerations make it likely that the presumed nonresponsiveness to mucosal flora may be more apparent than real in that normal organisms are responding to antigens in the mucosal microflora but only in the negative sense of inducing regulatory cells or the production of organism-specific IgA antibodies that regulate colonization and translocation. A final point of some interest is that the bacterial flora present in a particular niche in the intestine may have increased importance in eliciting inflammation in an animal model. Thus, creation of a cecal self-filling blind loop in HLA-B27 transgenic rats leads to proliferation of anaerobic bacteria, especially Bacteroides species and a more severe transmural cecal inflammation (103). Moreover, exclusion of the cecum leads to reduced gastric inflammation. Finally, in TCR-α chain–deficient mice, early removal of the tip of the cecum containing a large lymphoid aggregate leads to attenuation of subsequent colitis (107). Thus, it is possible that bacteria occupying a particular area of the intestine are of increased importance in generating effector cells that ultimately cause disease in all parts of the intestine. In summary, there can be no question from the foregoing discussion that the mucosal microflora play a critical role in models of mucosal inflammation by providing the major stimulus for the induction of effector T cells that cause the inflammation. This being said, it is also apparent that no single bacterial antigen has yet been shown to be responsible for this stimulation, although clearly some bacteria may be more important than others in this respect. Thus, whereas a continued search for a particular organism and/or antigen that causes IBD remains an important goal of some IBD investigators, the advent of models of mucosal inflammation that collectively show that mucosal inflammation is associated with

4 Feb 2002

11:44

510

AR

AR152-17.tex

STROBER

¥

FUSS

AR152-17.SGM

¥

LaTeX2e(2001/05/10)

P1: GKD

BLUMBERG

inherent immune defects in the face of an unaltered flora indicates that this goal may prove futile.

Annu. Rev. Immunol. 2002.20:495-549. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

Genetic Factors in Models of Mucosal Inflammation STUDIES IN MOUSE MODELS The possible genetic factors underlying models of mucosal inflammation have only recently received attention. Such factors could conceivably be operative both in models of spontaneous or induced colitis in mice strains with no known underlying genetic defects, or in mice with a known gene deletion or over-expression. In the latter case the genetic factor could conceivably influence the expression of the known gene defect. Evidence of such factors has been shown in relation to DSS colitis, in which it has been demonstrated that different mouse strains have different susceptibilities to disease (108, 109). Not surprisingly, strains in which mucosal inflammation has occurred spontaneously, such as the C3H/HeJBir mouse, have proved highly susceptible to DSS colitis, as did autoimmune-prone NOD mice of various types (108). Interestingly, Non-/LtJ mice were quite resistant to DSS colitis even though NON mice are congenic with NOD with respect to MHC. This lack of involvement of MHC genes in this form of colitis, if generalizable, is consistent with the view that no single antigen or set of antigens is involved in inducing experimental mucosal inflammation. In a second published study of genetic factors, a genome-wide search for quantitative trait loci (QTL) for susceptibility to DSS colitis in susceptible C3H/HeJ mice was conducted (110). In this study the C3H/HeJ mice were crossed with partially resistant C57BL/6 mice and strain-specific genetic areas associated with occurrence of colitis in their F2 progeny was determined. A number of QTLs were identified including those on chromosomes 1, 2, 5, and 18. In addition, several resistance loci were identified in susceptible NOD/Lt strain mice carrying resistance alleles from either B6 on chromosome 2 or from NON/Lt on chromosome 9. Thus, the genetic factors present in DSS colitis were highly complex. A third study of genetic factors in mouse models examined genetic factors controlling disease severity in IL-10–deficient mice (111). Here, IL-10–deficient mice on a C3H/HeJBir background manifested severe colitis when intercrossed with IL10–deficient mice on a C57BL/6 background that manifested mild colitis; this was done to determine inheritance of disease in the F2 generation and thus to identify QTLs. A C3H-derived colitogenic locus was found on chromosome 3 in two separate studies. This locus interacts in a complex fashion with other loci including a BL6-derived QTL on chromosome 18, a C3H-derived QTL on chromosome 8 for cecal lesions, and a C3H-derived disease QTL on chromosome 3, chromosome 9, and chromosome 19. These crosses thus found colitogenic susceptibility modifier genes that interact with IL-10 deficiency to cause more severe disease. Finally, an as yet unpublished study of genetic factors in TNBS colitis using susceptible SJL/J mice and resistant C57BL/6 mice subjected to a similar genomewide search revealed loci on chromosomes 9 and 11 (G. Bouma & W. Strober, unpublished observations). These findings parallel those derived from a recent

4 Feb 2002

11:44

AR

AR152-17.tex

AR152-17.SGM

LaTeX2e(2001/05/10)

P1: GKD

INSIGHTS INTO IBD

511

Annu. Rev. Immunol. 2002.20:495-549. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

study of SAMP1/Yit mice that also revealed the existence of a locus on chromosome 9 (112). Thus, a QTL in chromosome 9 may well harbor an important gene involved in susceptibility to colitis both in mouse models of inflammation and in humans. The QTL in chromosome 11 in the study of TNBS colitis is also of interest because the tendency of this strain of mice to manifest high IL-12 responses maps to the same region. Thus, the possibility emerges that a gene controlling IL-12 responses is an important susceptibility gene in a model of colitis as well as in IBD. STUDIES IN HUMANS WITH IBD One area of research into mucosal inflammation in which studies of humans with IBD can inform us about murine models of inflammation rather than vice versa is the area of genetic factors operating in these abnormalities. Two developments in the study of IBD are relevant. The first is that large-scale genome-wide searches conducted in families containing multiple members with IBD have led to the identification of 12 chromosomal loci associated with the occurrence of disease (113). In some but not all cases these loci have been confirmed by two or more independent studies and thus are genomic areas where disease genes can eventually be found. The finding that human IBD is a multigenic disease as implied in these human studies has relevance to the murine models, as it indicates that multiple genes are involved in the murine models even when the latter is due to a known genetic defect. This explains the fact that the expression of disease in, for example, Gi2a-deficient mice varies greatly with the strain of mouse bearing this defect. Finally, it is important to note that studies of susceptibility and resistance genes for murine mucosal inflammation, such as those discussed above, can greatly facilitate this search for disease genes in humans because the location of identified murine genes can ultimately be linked to syntenic genes in humans. The second development in the study of IBD is that the gene located in the most well established of the above loci, that in IBD-1, has recently been identified by two independent groups using two independent techniques (114, 115). These groups have shown that a gene encoding the protein present in macrophages and known as NOD-2 is a disease gene in Crohn’s disease; some 10–20% of individuals with the disease have mutations in NOD-2 and those that are homozygous for a mutated gene will invariably develop the disease. The function of the NOD-2 gene is poorly understand and thus its relation to the pathogenesis of IBD is essentially unknown. As reviewed by Beutler (116), some hints as to its function come from a knowledge of its structure: NOD-2 contains on one end a leucine-rich region where most of the mutations have been found and on the other end a caspase recruitment domain. Leucine-rich regions are thought to be binding regions and are found in toll-like receptors (TLRs). Thus, one possibility for the function of NOD-2 is that it interacts with ligands of TLRs (LPS and other bacterial products) that have gained entrance to the interior of the cell and then activate the NF-κB pathway through RICK, a protein known to bind to NOD-1, a close homologue of NOD-2. In this scenario, mutations of NOD-2 in Crohn’s disease are gain-of-function mutations that lead to increased NF-κB activation and inflammation. This would imply that

4 Feb 2002

11:44

Annu. Rev. Immunol. 2002.20:495-549. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

512

AR

AR152-17.tex

STROBER

¥

FUSS

AR152-17.SGM

¥

LaTeX2e(2001/05/10)

P1: GKD

BLUMBERG

processes involved in natural immunity (i.e., interactions of bacterial products with substances similar to those on TLRs) rather than adoptive immunity may play an important role in the initiation of Crohn’s disease. An alternative function of NOD-2 relates to the presence of the caspase recruitment domain and the possibility that NOD-2 is involved in caspase activation and apoptosis.One could propose that the mutations are loss-of-function mutations that lead to decreased apoptosis and thus to the persistence of cells that produce inflammatory cytokines. At the moment, neither of these possibilities is supported by either in vitro or in vivo evidence, and further studies are necessary to decide their validity or, indeed, the validity of other possibilities. Finally, with respect to the murine models of mucosal inflammation, none have yet been identified that appear to have NOD-2 mutations. Nevertheless, it is still possible that one or more of the models with no identifiable cause such as a SAMP1/Yit model can be due to a NOD-2 mutation, particularly because this model so closely resembles human Crohn’s disease (see discussion below). In addition, it is possible that, as implied above, a NOD-2 mutation acts as a contributing “background” abnormality determining susceptibility to mucosal inflammation primarily owing to another defect.

Th1 MODELS OF MUCOSAL INFLAMMATION By far the most common immunologic mechanisms leading to a model of mucosal inflammation are those involving a dysregulation of the Th1 T cell pathway. As already mentioned, the most important reason for this Th1 bias is that conditions in the mucosal environment, particularly the ubiquity of substances that induce IL-12, favor excessive Th1 response over excessive Th2 response if and when there is an imbalance in mucosal immune homeostasis. In general, the nature of the inflammation at both the macroscopic and microscopic levels in Th1 models is most closely related to Crohn’s disease, and indeed, this disease has quite clearly been shown to be due to a Th1 T cell disturbance (or a set of disturbances). Of the various Th1 models, two have been studied most intensively and have yielded insightful information. We discuss these models in some detail.

TNBS Colitis Hapten-induced colitis [trinitrobenzene sulfonic acid (TNBS)–colitis] is an important model of mucosal inflammation because it allows for the study of early or initiating events in the development of a mucosal inflammation and because it allows analysis of the relation of the response to a specific antigen (a hapten) to the overall mucosal immune response leading to colonic inflammation. Whereas this model has been the object of study for over two decades (117, 118), it was not until 1995 when Neurath et al. showed that TNBS administered per rectum (in the presence of ethanol) to SJL/J mice resulted in a transmural infiltrative

4 Feb 2002

11:44

AR

AR152-17.tex

AR152-17.SGM

LaTeX2e(2001/05/10)

P1: GKD

Annu. Rev. Immunol. 2002.20:495-549. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

INSIGHTS INTO IBD

513

disease limited to the colon and owing to an IL-12–driven, Th1-mediated response (32). The latter was definitively demonstrated by the fact that treatment of mice with only a single dose of anti–IL-12 antibody results in complete prevention of TNBS colitis or, in the case of mice with preexisting and ongoing disease, in complete and rapid disappearance of the inflammatory lesion (32). In addition, it was shown that, in common with the colonic inflammation seen in IL-2-deficient mice or in several other models of colonic inflammation, the disease could be prevented by administration of anti-CD40L (CD154), indicating that the Th1 response driving this Th1-mediated inflammation was based on CD40L-CD40 interactions (91, 119–121). In subsequent studies of the role of the various IL-12–induced Th1 cytokines participating in the pathogenesis of TNBS colitis, it has been shown that the role of TNF-α is surprisingly important. In particular, TNBS could not be induced in TNF-α–deficient mice and is far more severe in mice that over-express this inflammatory cytokine (122). One possible explanation of these findings is that TNF-α is necessary for both the initiation and persistence of the Th1 response, possibly by acting as a proximal cofactor for IL-12 or IL-18 production. The dramatic effect of anti–IL-12 antibody administration on TNBS colitis (and as subsequently shown, on other murine models of colitis) can be linked to the observation that such administration is associated with increased numbers of TUNEL-positive cells in lamina propria tissues and in dispersed cell populations (123). This, plus administration of Fas-Fc to mice undergoing treatment with anti–IL-12, strongly supports the idea that anti–IL-12 treatment leads to Fasmediated Th1 T cell apoptosis and that TNBS colitis is rapidly responsive to anti– IL-12 because the latter leads to the death of the Th1 T cells inducing the colitis. One of the major insights derived from the study of TNBS colitis is that regulatory mechanisms inherent in the mucosal immune responses can prevent the development of colitis. This was shown by the fact that whereas intrarectal administration of TNBS led to colonic disease, oral administration of TNBS in the form of TNP-haptenated colonic protein (TNP-CP) prevented colitis induced by intrarectal TNBS administration (124, 125). In addition, it was shown that the preventive effect was due to the induction of regulatory cells producing TGF-β, because TNP-CP feeding led to the appearance of TGF-β-producing cells in the lamina propria, and coadministration of anti–TGF-β antibody to mice fed TNP-CP abrogated the protective effect. This TGF-β -mediated protection is due to the induction of oral tolerance in the face of an induced mucosal inflammation and indicates that immune responses resulting in inflammation of the mucosa are as subject to mucosal regulatory effects as the response resulting from the feeding of protein antigens (24, 67–69). On this basis, it is reasonable to attribute the induction of TNBS colitis in SJL/J mice by intrarectal TNBS administration alone to the fact that such administration engenders a mucosal Th1 T cell response that is not balanced by the prompt appearance of a regulatory response. We return to the possible reason why this is so after we discuss the role of mucosal microflora in TNBS colitis.

4 Feb 2002

11:44

Annu. Rev. Immunol. 2002.20:495-549. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

514

AR

AR152-17.tex

STROBER

¥

FUSS

AR152-17.SGM

¥

LaTeX2e(2001/05/10)

P1: GKD

BLUMBERG

In other models of mucosal inflammation, most notably that seen in IL-10– deficient mice, IL-10 rather than TGF-β appears to be the major cytokine-mediating regulation. In recent, as yet unpublished, studies of the relation between TGF-β and IL-10 production in TNBS colitis involving administration of anti–TGF-β and anti–IL-10 as well as adoptive transfer of regulatory cell populations, it was shown that TGF-appears to be the major regulatory cytokine but that IL-10 is necessary for the maintenance and/or the effectiveness of the TGF-β response (125a). Thus, in the key experiment of these studies it was shown that whereas mice fed TNP-CP and then given TNBS per rectum to induce colitis were protected from colitis, administration of anti–IL-10 after feedings prevented protection and reduced both IL-10 and TGF-β responses; however, administration of anti–TGF-β prevented protection and reduced TGF-β responses but not IL-10 response. Thus, TGF-β levels were more closely associated with counter-regulation than were IL-10 levels. In recent studies taking advantage of the potent capacity of TGF-β regulatory cells to ameliorate TNBS colitis, DNA encoding active TGF-β was administered to mice intranasally to induce genetically engineered T cells producing TGF-β in vivo (29). Indeed, following such treatment, T cells and macrophages producing TGF-β were subsequently found in lamina propria and spleen, where they acted to prevent induction of and treat TNBS colitis. Interestingly, the induction of such regulatory cells was associated with production of high levels of IL-10, which also contributed to the regulatory effect. These studies open the door to the possibility that gene therapy with genes encoding regulatory cytokines will become a viable form of treatment of Th1 mucosal inflammation. Whereas TNBS (or more specifically the TNP epitope) may be the main antigenic stimulus that drives the Th1 responses in TNBS colitis, it is likely that other antigenic determinants present in the mucosal microflora also contribute to the immune response driving this disease. This view stems from evidence reviewed above, which shows that mice with TNBS colitis react to their own microflora and that such reactivity disappears with anti–IL-12 treatment (87, 88, 90). Reactivity to mucosal microflora also relates to TNBS colitis in a way that bears on genetic factors in this model. Recall that TNBS administered per rectum to induce colitis is administered with ethanol, a substance that disrupts the mucosal barrier and thus, as an initial event, causes increased exposure of the mucosal immune system to mucosal microflora. In SJL/J mice that are susceptible to colitis there is evidence that such exposure leads to a high IL-12 response, and it is reasonable to suppose that this sets in motion a massive Th1 response to TNBS that precludes a concomitant regulatory TGF-β response. In contrast, administration of TNP-CP by mouth (in the absence of ethanol) does not lead to an initial IL-12 response, and thus a normal mucosal response replete with a regulatory component ensues. Carrying this concept one step further, one might postulate that mouse strains susceptible to TNBS colitis are precisely those that mount high IL-12 when exposed to mucosal microflora. Evidence that this is the case comes from the study of susceptibility loci in SJL/J mice mentioned above, showing that two such loci controlling TNBS

4 Feb 2002

11:44

AR

AR152-17.tex

AR152-17.SGM

LaTeX2e(2001/05/10)

P1: GKD

INSIGHTS INTO IBD

515

Annu. Rev. Immunol. 2002.20:495-549. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

colitis can be identified, and one of these is similar if not identical to that associated with high IL-12 responsiveness (G. Bouma, W. Strober, submitted for publication). In summary, TNBS colitis in SJL/J mice is an immunologically mediated colitis that results from the rapid induction of an IL-12–driven, Th1-mediated response that precludes development of a counter-regulatory TGF-β response. As such, this is a type 1 model in that the major driving force is the overactivity of disease-causing effector cells. Emerging evidence suggests that this response is genetically controlled, most probably by genes that regulate the magnitude of an initial IL-12 response to substances in the mucosal microflora.

The SCID-Transfer Model of Colitis A second important model of mucosal inflammation is that produced by repletion of SCID or Rag2−/− with either CD45RBhi T cells (naive T cells) or with a combination of CD45RBhi T cells and CD45RBlo T cells (mature T cells) (73, 126, 127). In the former case repletion leads in 3–5 weeks to severe colitis, whereas in the latter case no inflammation occurs. Herein lies the power of the model: One can immediately identify two cell populations, one a source of effector cells and the other a source of regulatory cells, and one can conduct analyses of each population to identify the cells necessary for each type of function. In initial studies of this model it was found that the inflammation was due to a Th1-mediated T cell response, also driven by IL-12 and mediated by IFN-γ (128, 129). In this instance whereas the colitis was less effectively inhibited by anti–TNF-α treatment than TNBS colitis, cells from STAT4-deficient mice still gave rise to disease, perhaps because of their continued ability to produce TNF-α (130). Whereas CD45RBhi cells populate the small intestine as well as the large intestine of SCID recipients undergoing cell transfer, inflammation is limited to the colon. This immediately suggested that organisms endogenous to the colon provide the antigenic stimulus for the mucosal inflammation in this model. Evidence in support of this concept came from studies of the SCID-transfer model that showed that transfer of cells to mice reared in a “near gnotobiotic” environment (rather than the specific pathogen-free environment of the mice in the original studies) manifested greatly reduced levels of inflammation (94). In addition, CD4+ T cells from colitic mice proliferated and produced Th1 cytokines in response to antigen presenting cells pulsed with fecal extracts of normal but not germ-free mice (94). If indeed mucosal microflora drive effector cells in SCID-transfer colitis, the number of stimulating antigens in the microflora is circumscribed, because colitic mice contain populations of cells with restricted T cell receptor (TCR) diversity and expression of particular CDR3 sequences (131). It should be noted, however, that the selected TCRs differ from mouse to mouse despite MHC class II identity, indicating that the number of potentially stimulating antigens may be considerable. Finally, the expanded clones were widely dispersed in lymphoid tissue and were

4 Feb 2002

11:44

Annu. Rev. Immunol. 2002.20:495-549. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

516

AR

AR152-17.tex

STROBER

¥

FUSS

AR152-17.SGM

¥

LaTeX2e(2001/05/10)

P1: GKD

BLUMBERG

detected early (prior to development of overt colitis), indicating that stimulation of these clones was an early event in disease pathogenesis. Overall, these data are compatible with the view that cell activation in SCID-transfer colitis is an antigendriven event occurring as a result of inappropriate responsiveness to antigens in the mucosal microflora. As noted above, the SCID-transfer model is particularly useful for the study of regulatory cells in mucosal inflammation. In an early study of such regulation it was shown that the protective effect of CD45RBlo T cells was not due to the secretion of IL-4 because the cells that mediated the protection could be obtained from IL-4–deficient mice and protection occurred in spite of repeated administration of anti–IL-4; however, it was due to secretion of TGF-β because in this case repeated anti–TGF-β administration reversed protection (132). In more recent studies of the relationship of TGF-β and the regulation of SCID-transfer colitis it was shown that the regulatory cells were CD25+ T cells because CD45RBlo cells depleted of CD25+ T cells were unable to prevent colitis (133). In addition, evidence was presented that the protection afforded by CD25+ T cells was abolished by co-administration of anti–CTLA-4 and anti–TGF-β antibodies, indicating that this subset required stimulation via CTLA-4 and either produced TGF-β itself or induced such secretion in other cells. That the former possibility is correct is supported by recent studies by Nakamura et al., who showed that CD25+ T cells produce TGF-β when stimulated by anti-CD3 antibody under crosslinking conditions and costimulated with anti-CD28 or anti–CTLA-4 (59). Thus, the picture that emerges is that CD25+ T cells in the CD45RBlo T cell populations secrete TGF-β in a CTLA-4–dependent fashion to mediate suppression of colitis. Parenthetically, recent evidence suggests that CD25+ cells express surface TGF-β in the form of a latent (inactive) protein associated with latency-associated protein. This surface TGF-β may be responsible for CD25+ T cell suppression mediated by cell-cell contact, i.e., the form of suppression exerted under suboptimal stimulation and in the absence of overt inflammation. A final point to be made about the regulation of SCID-transfer colitis by CD25+ T cells relates to the fact that, as shown originally by Sakaguchi et al. (134), such cells develop in the thymus and, as recently shown by Bensinger et al. (135), are dependent on the presence of MHC class II–positive cortical epithelial cells for their intrathymic development. Thus, it is not surprising that CD25+ cells from MHC class II–deficient mice neither act as suppressor cells in in vitro assays nor suppress colitis when injected together with CD4+ CD45RBhi cells into Rag-2– deficient recipients. This evidence that CD25+ cells regulating colitis can originate in the thymus should not be taken to mean that this is the only site of development of these regulatory cells. It remains possible (albeit unproven) that such cells also develop, or at least undergo expansion, in the mucosal tissues. IL-10, no less than TGF-β, has also been implicated in the regulation of SCID colitis. Initial evidence for this came from a study showing that CD45RBhi T cells do not cause colitis if obtained from IL-10 transgenic mice (136). In further studies, Groux et al. showed first that T cell clones expanded in vitro in the presence

4 Feb 2002

11:44

AR

AR152-17.tex

AR152-17.SGM

LaTeX2e(2001/05/10)

P1: GKD

Annu. Rev. Immunol. 2002.20:495-549. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

INSIGHTS INTO IBD

517

of IL-10 produce high levels of IL-10 (and IL-5) and, in some cases, TGF-β as well (55). They then showed that these clones suppress T cell response in vitro and, more importantly, suppress SCID-transfer colitis when administered in place of CD45RBlo T cells. These studies thus defined a new class of regulatory cells (called Tr1 cells) that mediate immune suppression mainly via IL-10. It should be noted, however, that these cells also produce TGF-β and thus the relevant suppressor cytokine is uncertain. A final series of studies relative to IL-10 and regulation of SCID-transfer colitis showed that CD45RBlo T cells obtained from IL-10– deficient mice do not prevent colitis when administered with CD45RBhi T cells; similarly, treatment of SCID mice administered both CD45RBhi and CD45RBlo T cells with anti–IL-10 receptor led to the development of colitis (137). These studies thus show that IL-10 is necessary for protection against colitis even if it is not sufficient. Taken together, the above studies show quite definitively that IL-10 and TGF-β are important regulatory cytokines in SCID-transfer colitis. The question arises here, even more than in the case of TNBS colitis, as to how these regulatory cytokines interact to bring about regulation. To date, no studies addressing this question in the context of SCID-transfer colitis have appeared; however, based on the data derived from the TNBS colitis model we would suggest that TGF-β is the major suppressor cytokine and that IL-10 is needed to facilitate TGF-β secretion and/or activity. Yet another cell type contributing to regulation of SCID-transfer colitis is the NK cell. This is supported by the fact that transfer of CD45RBlo T cells into NK cell–depleted recipients results in more severe colitis (60). Such regulatory effects, as noted above, are also seen in other models of mucosal inflammation and occur via an unknown mechanism. However, it is know that this form of regulation is distinct from that mediated by CD45RBlo cells inasmuch as depletion of the latter of NK cells does not eliminate their regulatory effect. One possible but highly speculative explanation is that NK cells lyse activated effector cells or APCs inducing the effector cells. This concept is supported by the fact that NK cells from perforin-deficient mice have no regulatory effects (60). A final point to emerge from studies of the SCID-transfer model relates to the recent observation that colitic mice have greatly increased numbers of CD134+ (OX40+) dendritic cells in their mesenteric lymph nodes and that administration of anti-CD134L antibody leads to reversal of colitis (as it does in other models) (43, 138). It is likely that these mesenteric lymph node cells originate in the inflamed lamina propria and then migrate to the draining lymph node where they provide inductive signals to effector T cells about to migrate into lesional tissues. This would imply that much of the antigen presentation necessary for the development of mucosal lesions goes on in regional lymph nodes, rather than in the inflamed tissue itself. In summary, the SCID-transfer model is a model of mucosal inflammation that allows one to separate effector and regulatory T cell functions mediating the inflammatory process. Thus, by analysis of both effector cells in CD45RBhi

4 Feb 2002

11:44

518

AR

AR152-17.tex

STROBER

¥

FUSS

AR152-17.SGM

¥

LaTeX2e(2001/05/10)

P1: GKD

BLUMBERG

Annu. Rev. Immunol. 2002.20:495-549. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

T cell populations and regulatory cells in CD45RBlo T cell populations, it has been possible to clearly demonstrate the interplay between various cell types that determine whether mucosal inflammation will occur. In addition, this model demonstrates that abnormal reactivity to antigen in the mucosal microflora can develop in the absence of a genetic abnormality and does not require initial disruption of the epithelial cell barrier. Rather, the only precondition for the occurrence of colitis is a marked imbalance between effector and regulatory cell populations.

Type 1 Models of Colitis: Defects that Directly or Indirectly Affect the Synthesis of Key Cytokines in the Th1 Pathway of T Cell Differentiation Joining TNBS colitis as type 1 defects leading to mucosal inflammation are several models whose pathogenesis can be traced to abnormalities that lead to the overproduction of key cytokines in the Th1 T cell differentiation pathway, such as IFN-γ , TNF-α, and IL-12. COLITIS ASSOCIATED WITH DEFECTS IN THE PRODUCTION OF TRANSCRIPTION FACTORS CONTROLLING IFN-γ PRODUCTION Given the central role of IFN-γ in the

Th1 responses, it should come as no surprise that molecular defects resulting in IFN-γ overproduction can lead to colitis. Two such defects are now known to exist, one affecting T-bet and one affecting STAT4. As shown by Szabo et al., T-bet is a T-box protein that when over-expressed in T cells programs them for high IFN-γ responses and low IL-4 responses, even when the cell is a supposedly “committed” Th2 cell (139). This, plus recent data showing that over-expression of T-bet can lead to IFN-γ responses in cells lacking STAT4, has led to the concept that this factor is the molecular switch for Th1 differentiation (140). STAT4, on the other hand, has also been shown to be a necessary factor for Th1 differentiation that probably acts as both a transcription factor for IFN-γ and as a factor that maintains Th1 T cell survival (141, 142). In the relevant studies of colitis associated with a STAT4 abnormality, it has been shown that mice bearing a STAT4 transgene (under a CMV promotor) develop colitis when administered TNP-KLH in Freund’s adjuvant, an antigenic stimulus that has no colitogenic effect in normal mice (72). In addition, spleen cells from these mice proliferate when exposed to antigens in their autologous microflora in vitro, and T cells thus stimulated induce colitis in SCID recipients. Corresponding studies of T-bet abnormalities have shown that naive T cells from T-bet–deficient mice exhibit a reduced capacity to transfer colitis to SCID mice, whereas, conversely, naive T cells over-expressing T-bet (owing to infection with a T-bet-expressing retrovirus) induce accelerated colitis in SCID mice (M. Neurath, R. Blumberg, L. Glimcher, manuscript submitted). In addition, memory T cells from T-bet–deficient mice exhibit an enhanced capacity to protect SCID mice from colitis when cocultured with naive T cells.

4 Feb 2002

11:44

AR

AR152-17.tex

AR152-17.SGM

LaTeX2e(2001/05/10)

P1: GKD

Annu. Rev. Immunol. 2002.20:495-549. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

INSIGHTS INTO IBD

519

MUCOSAL INFLAMMATION IN MICE THAT OVER-EXPRESS TNF-α: TNF1ARE MICE Also in the category of models of inflammation owing to abnormalities of Th1 cytokine production is the model owing to over-expression of TNF-α, the TNF1ARE mouse (36). This model results from a targeted deletion of AU-rich elements located in the 30 untranslated region of the TNF-α gene, which gives rise to dysregulation of the processing of TNF-α mRNA and the overproduction of TNF-α protein. The phenotype of these mice is notable because the mucosal inflammation is mainly located in the terminal ileum and only occasionally in the proximal colon; in addition, it is remarkably similar to that in Crohn’s disease: It is a transmural infiltrative lesion that contains typical granulomata. Of considerable interest, TNF1ARE mice develop arthritis resembling rheumatoid arthritis, as well as mucosal inflammation. Studies of the immunopathogenesis of disease in TNF1ARE mice indicate that whereas the mucosal inflammation is dependent on the presence of T cells, the joint inflammation is not (36). This, plus the fact that the different TNF receptors are involved in the two kinds of inflammations, indicates that the pathogenesis of inflammation in the two areas is different. As far as the disease in the mucosa is concerned, it is likely that the inflammation is initiated and maintained by substances in the mucosal microflora that can induce TNF-α production (LPS, CpGs, etc.). However, it is important to mention that colitis does not occur in TNF1ARE mice that are also IL-12 p40–deficient (F. Cominelli, personal communication); thus, the inductive process does not appear to involve the direct stimulation of cells by microfloral stimuli, but rather their indirect stimulation via IL-12. It is important to mention here that IL-12 production may be enhanced in TNF1ARE mice because once TNF-α overexpression is initiated, a positive feedback loop between IL-12 and TNF-α is established.

COLITIS IN Gi2A MICE Mice deficient in the G protein Gi2α provide yet another type 1 model in which overproduction of a Th1 cytokine results in colitis. The inflammation in this model is a Th1-mediated colitis with an infiltrative histologic picture similar to other Th1 colitides (143). The basis of this colitis has been elucidated by studies showing that a stimulus that inhibits Gi protein signaling, such as pertussis toxin, enhances splenocyte production of IL-12, TNF-α, and IL10 in vitro upon culture with Staphylococcus aureus Cowan I and CD40L (144). In addition, pertussis toxin–treated BALB/c mice exhibit a healing phenotype when infected with Leishmania major, whereas untreated mice of this strain manifest progressive infection. That these findings are relevant to Gi2α mice was shown by the fact that these mice produce increased amounts of IL-12 and TNF-α when their CD8a+ (lymphoid) dendritic cells are appropriately stimulated. Thus, it is reasonable to conclude that the colitis in Gi2α-deficient mice is a type 1 colitis owing to the overproduction of IL-12. One can reasonably assume that antigens in the mucosal microflora are the driving force of such overproduction, but this needs to be verified experimentally.

4 Feb 2002

11:44

520

AR

AR152-17.tex

STROBER

¥

FUSS

AR152-17.SGM

¥

LaTeX2e(2001/05/10)

P1: GKD

BLUMBERG

Type 2 Models of Colitis: Defects in the Production of Proteins that are Directly or Indirectly Involved in Regulation of Mucosal Responses

Annu. Rev. Immunol. 2002.20:495-549. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

The mirror image of the previous category of models of mucosal inflammation are those in which there is an abnormality in the synthesis of a regulatory cytokine or a protein that affects the function of a regulatory cytokine. These are thus type 2 models and include colitis owing to IL-10 deficiency or abnormalities of IL-10 signaling and to defects in TGF-β function. COLITIS OWING TO IL-10 DEFICIENCY OR IL-10 SIGNALING DEFECTS As initially reported by Kuhn et al., IL-10–deficient mice raised in a specific pathogen-free (SPF) or a conventional environment develop colitis marked by epithelial cell hyperplasia and a transmural inflammation (145). Early on, the disease is due to a Th1 response and is completely ameliorated by anti–IL-12 treatment. Later, however, a Th2 response supervenes and the lesion is no longer treatable with anti–IL-12 (35; A.D. Levine, personal communication). The reason for this change is unclear, but it may relate to the fact that in the absence of IL-10 down-regulation, Th2 responses ultimately prevail. Further studies of the cytokines involved in the colitis of IL-10 deficiency have come from studies in which colitis was induced in SPF mice by infection with H. hepaticus. It was shown that whereas IL-12 had to be present for colitis to develop (99), IFN-γ or TNF-α did not, as treatment with anti–IFN-γ or anti–TNF-α had no effect on the colitis; it is thus apparent that each of these cytokines can mediate colitis in the absence of the other (146). Additional studies showed that treatment with both anti–IFN-γ and anti–TNF-α was also not effective in the treatment of colitis, suggesting that in IL-10 deficiency yet other cytokines induced by IL-12 may play effector roles in the absence of both IFN-γ and TNF-α. As in other models, the colitis of IL-10–deficient mice does not develop under germ-free conditions and is thus driven by antigens in the mucosal microflora (93, 99, 100). Recent studies of colitis in SPF mice infected with H. hepaticus underscore this fact. Thus, the study mentioned above showed that such infection led to greatly enhanced colitis, but another study showed that such infection led to no more colitis than that ordinarily seen under SPF conditions (99, 100, 146). This suggests that mucosal microflora (such as that present in some mouse holding areas) can influence the development of inflammation even when it is driven by a known pathogen. A final point concerning the role of the mucosal microflora in the colitis of IL-10–deficient mice is that these mice manifest increased intestinal permeability even prior to the development of overt colitis. This change in barrier function may lead to increased contact with or stimulation by antigens in the mucosal microflora and is thus a factor that facilitates the development of the inflammation (66). The basis of the immunoregulatory defect leading to colitis in IL-10–deficient mice undoubtedly lies in the fact that IL-10 has major suppressive effects on

4 Feb 2002

11:44

AR

AR152-17.tex

AR152-17.SGM

LaTeX2e(2001/05/10)

P1: GKD

Annu. Rev. Immunol. 2002.20:495-549. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

INSIGHTS INTO IBD

521

immune responses, and its absence in an area of the body constantly exposed to antigens leads to inflammation in that area. The negative effects of IL-10 on immune responses is well demonstrated in numerous in vitro studies showing that IL-10 inhibits IL-12 and TNF-α production, suppresses costimulatory molecules, and directly inhibits T cell proliferation and/or induces T cell apoptosis (147–150). In addition, in vivo studies of IL-10 transgenic mice and the SCID-transfer model discussed above have shown that mature CD45RBlo T cells from IL-10 deficient mice are incapable of preventing the development of colitis (137). This finding is underscored by the aforementioned studies in which Tr1 cells that produce IL-10 can substitute for CD45RBlo T cells in the prevention of SCID-transfer colitis (55). Taken together, these data provide ample reason to postulate that the colitis of IL-10 deficiency is a prototypic type 2 model of colitis owing to absence of a major regulatory cytokine. The only question that remains with respect to this conclusion is the one discussed previously in relation to TNBS-colitis concerning the relation of TGF-β and IL-10 in the regulation of mucosal inflammation. It was mentioned in that context that the evidence now available favors the view that TGF-β is the more proximal cytokine suppressor and that the main role of IL-10 is to maintain and facilitate the TGF-β suppressor effect. This leads to the supposition that in the colitis of IL-10 deficiency it is not the lack of IL-10 per se that leads to inflammation, but rather the lack of an adequate TGF-β response that occurs in the absence of IL-10. Finally, it is important to note that not only IL-10 deficiency can lead to colitis, but so can defects in IL-10 signaling that functionally are equivalent to IL-10 deficiency. This is seen in mice deficient in an “orphan” receptor termed CRF-2, which forms part of the IL-10 receptor, and in mice whose macrophages and neutrophils are deficient in STAT3 expression that exhibit a defect in IL-10 signaling (151, 152). COLITIS ASSOCIATED WITH TGF-β DEFECTS In light of the discussion above it is to be expected that deficiency in the production of TGF-β should also lead to type 2 models of mucosal inflammation. This was presaged by studies of mice with TGF-β1 deficiency owing to targeting of the TGF-β1 gene, in which it was shown that such mice exhibit widespread inflammation in multiple organs and early death (153, 154). It should be noted, however, that this inflammation is not more prominent in mucosal tissues than in other tissues, possibly because the mice die of widespread autoimmune disease before they have the chance to develop inflammation to “exogenous” mucosal antigens. A more particular relation of defects in TGF-β function to mucosal inflammation occurs in mice with defective TGF-β signaling, either only in T cells or only in epithelial cells owing to the presence of transgenes encoding dominant-negative TGF-β receptors (TGF-βRII) under T cell–specific or epithelial cell–specific promotors, respectively (75, 155). In the former case inflammation develops in the colon and lung and the mice develop autoantibodies and glomerular immune complex deposition. Interestingly, both Th1 and Th2 cytokine production is increased,

4 Feb 2002

11:44

522

AR

AR152-17.tex

STROBER

¥

FUSS

AR152-17.SGM

¥

LaTeX2e(2001/05/10)

P1: GKD

BLUMBERG

probably because TGF-β regulates both Th1 and Th2 responses. In the latter case of the epithelial-specific expression of the dominant-negative transgene, the mice develop colitis under conventional conditions and manifest increased susceptibility to the development of dextran sulfate colitis. The development of colitis in these mice suggests that TGF-β also regulates epithelial cell function and in its absence the mucosal is more subject to stimulation by antigens in the mucosal flora.

Annu. Rev. Immunol. 2002.20:495-549. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

Colitis in which the Presence of Type 2 Defects Are Traceable to Thymic Dysfunction Whereas mice with IL-2 deficiency (or IL-2R deficiency) obviously have a very different immunologic defect than bone marrow–reconstituted Tgε26 mice, the pathogenesis of the colitis associated with these defects is sufficiently similar to warrant their discussion under the same heading. Thus, in both models the inflammation is an IL-12–driven, Th1-mediated inflammation that depends on CD40L-CD40 interactions and is abrogated by the absence of IL-12 (89, 120, 121, 156). In addition, in both models there is evidence that the underlying abnormality is the defective generation of regulatory cells and that an intrathymic defect is probably responsible for this abnormality (157–159). COLITIS IN IL-2–DEFICIENT MICE Mice with IL-2 deficiency exhibit early development of lymphadenopathy, bone marrow infiltration, and hemolytic anemia indicative of a generalized autoimmune state, which are then followed in surviving mice by the development of a transmural colitis (86). The latter is a T cell–driven event that, as mentioned above, is due to a Th1 response (89). T cells in these mice bear markers of maturity and proliferation that may relate to the presence of an apoptosis defect, presumably occurring because in absence of IL-12 there is deficient activation-induced (Fas-mediated) apoptosis (86, 160). In the absence of IL-2, IL-15 may play a major role in this hyperproliferative state, but this is not supported by the relevant available in vitro studies (161). As in other models, it is clear that antigens in the mucosal microflora drive the T cells because disease does not develop under germ-free conditions (81, 86). In addition, there is some evidence that in the absence of IL-2, epithelial cells exert increased antigen-presenting function, which plays a role in T cell activation and cytokine secretion (162). The pathogenesis of colitis in IL-2–deficient mice has been successfully studied by inducing the rapid onset of colitis by intra-peritoneal injection of TNP-KLH in Freund’s adjuvant (89). This maneuver apparently stimulates T cells that crossreact with antigens in the mucosal microflora and mediate disease. Studies in this induced model of mucosal inflammation disclosed that whereas normal mice react to TNP-KLH stimulation with an IL-4/ TGF-β response, this response is absent in IL-2 deficient mice (89). Furthermore, they showed that concomitant administration of anti-CD3 does elicit an IL-4/TGF-β response and prevents development of disease. That such prevention is due to the TGF-β response and not the IL-4 response was shown by the fact that the protection obtained with anti-CD3 treatment was abrogated by simultaneous treatment with anti-TGF-β but not with

4 Feb 2002

11:44

AR

AR152-17.tex

AR152-17.SGM

LaTeX2e(2001/05/10)

P1: GKD

Annu. Rev. Immunol. 2002.20:495-549. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

INSIGHTS INTO IBD

523

anti–IL-4. Thus, these studies led to the conclusion that the underlying cause of colitis in IL-2 deficiency is a type 2 defect and an inadequate regulatory response (157). Further studies of immune function of IL-2–deficient mice as it relates to colitis focused on thymic function because it was known that thymic defects contribute to the autoimmunity seen in these mice (159). Again using the TNP-KLH to induce colitis in IL-2 deficient mice, it was shown that this stimulus leads to the appearance of increased numbers of single positive thymocytes in the thymus that displayed a Th1 cytokine secretion profile and transferred colitis to normal mice (159). These studies thus suggest that thymocyte development is defective in IL-2 deficiency and this defect leads to either increased numbers of effector cells capable of mediating either autoimmunity or colitis or to decreased numbers of regulatory cells capable of preventing these phenomena. Given the above role of regulatory cell dysfunction in IL-2–deficient mice, the latter rather than the former is likely to be the more important factor. Finally, it is important to mention that mice with IL-2R deficiency owing to either α or β chain gene targeting also develop autoimmunity and colitis (163–165). However, those with IL-2R deficiency owing to γ chain targeting do not, presumably because such mice cannot mount adequate T cell responses to support autoimmunity. IL-2Rβ chain–deficient mice differ somewhat from IL-2–deficient mice in that they display hypergranulopoiesis that crowds out normal marrow elements and leads to massive lymphoid infiltration with granulocytes (164, 165). In addition, they manifest poor responses to antigen, presumably because their cells respond poorly to both IL-2 and IL-15. This raises the question as to which cytokine is driving their autoimmune responses and leads to the possibility that in IL-2–deficient mice neither IL-2 nor IL-15 is necessary to support T cell responses. COLITIS IN BONE MARROW–RECONSTITUTED Tgε26 MICE Tgε26 mice are mice bearing a transgenic CD3-epsilon chain whose over-expression results in intrathymic T cell and NK cell death probably because of excessive signal transduction during thymic development (158, 166). In addition, they manifest a secondary defect in thymic stromal architecture because the development of the latter depends on the presence of normal thymocytes (167). Fetal mice bearing the transgene can be rescued by transplantation of T cell–depleted normal bone marrow because such transplantation preserves stromal architecture; in contrast, adult mice cannot be thus rescued because by this time the defect in architecture cannot be reversed (158, 167). Thus, whereas bone marrow reconstitution of adult mice leads to repair of lymphoid depletion, the reconstituted mice contain a cell population that has developed in a defective thymic micro-environment. Mice with the Tgε26 defect who are reconstituted with normal bone marrow (reconstituted Tge26 mice) develop an infiltrative colitis similar to that seen in IL-2 deficiency, which is due to an IL-12 driven Th1-mediated response driven by antigens in the mucosal microflora (158, 168). The IL-12 dependency of the inflammation is nicely shown by the fact that reconstitution of the mice with bone marrow from STAT4-deficient mice exhibit a greatly reduced level of disease, as do

4 Feb 2002

11:44

Annu. Rev. Immunol. 2002.20:495-549. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

524

AR

AR152-17.tex

STROBER

¥

FUSS

AR152-17.SGM

¥

LaTeX2e(2001/05/10)

P1: GKD

BLUMBERG

mice treated with anti–IL-12 (156). It should be noted, however, that reconstitution of mice with bone marrow from IFN-γ –deficient mice only slightly ameliorates the development of colitis, presumably because in this situation, as in the case of TNBS colitis and IL-10 deficiency colitis, TNF-α can induce inflammation in the absence of IFN-γ (156). In studies of the underlying factors leading to colitis in Tgε26 mice, cell transfer studies were performed using mice with and without transplantation of syngeneic normal fetal thymus (158). These studies showed first that transfer of nonmucosal cells from Tgε26 mice with colitis into untransplanted Tgε26 recipients resulted in colitis similar to that in the donor mice. It was then shown that Tgε26 mice reconstituted with normal bone marrow and transplanted with syngeneic fetal thymus (from Tgε26 mice) did not develop colitis, whereas those only reconstituted with bone marrow developed colitis. Because the transplanted fetal thymus maintained a normal architecture in mice reconstituted with normal marrow, the mice with the transplants could generate a cadre of normal cells that then intermixed with the abnormal cells arising from the abnormal thymus. Thus, these studies suggest that the abnormal cell population causes colitis because it lacks a regulatory cell population and that the colitis is a type 2 colitis. The above discussion makes it apparent that in IL-2 deficient mice and in bone marrow–reconstituted Tgε26 mice abnormal T cell development in the bone marrow is a major factor in the development of colitis. One theoretical difference, however, is that in the induced IL-2 deficiency model studied, the inducing antigen (TNP-KLH) is exogenous, whereas in reconstituted Tgε26 no exogenous antigens are introduced. This difference may be more apparent than real, however, because it is possible that mucosally derived antigens normally enter the thymus and affect thymic selection. Another difference is that in IL-2 deficiency the development of colitis may occur solely because of local mucosal dysregulation, whereas in reconstituted Tgε26 mice the thymus appears to be more intrinsic to the disease state.

Miscellaneous Models of Colitis: Th1 Responses Whose Underlying Immunopathogenesis is Not Understood Several models of mucosal inflammation have been described in which the basis of the mucosal inflammation has not yet been elucidated. In some cases these models may prove to be quite important because they may be due to one or more defects also present in humans with IBD. Two of these models are described in some detail below and the rest are summarized in Table 4. A substrain of LPS-unresponsive mice [lacking tolllike receptor 4 (TLR-4)] C3H/HeJ mice, termed C3H/HeJBir mice, have been noted to spontaneously develop a colitis centered in the cecum and proximal colon (169). The colitis consists of an transmural inflammation that begins at 3–6 weeks and gradually wanes. It is an IL-12–driven, Th1-mediated inflammation that can be transferred to SCID recipients by CD4+ T cells from the affected mice.

COLITIS IN C3H/HeJBir MICE

4 Feb 2002

11:44

AR

AR152-17.tex

AR152-17.SGM

LaTeX2e(2001/05/10)

P1: GKD

525

INSIGHTS INTO IBD TABLE 4 Miscellaneous Th1 and Th2 models of mucosal inflammation Salient/unique immunopathologic features

Annu. Rev. Immunol. 2002.20:495-549. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

Model

Type

Proposed mechanism

Reference

TCR transgenic mice with lymphopenia

Similar to SCID-transfer colitis; presence of Tg T cells that cross-react with mucosal antigens in most cases.

Th1, ? Type 2

Failure of intrathymic development of regulatory cells

204

IL-7 transgenic mice

Infiltrative lesion with crypt abscesses and loss of goblet cells; IL-7 over-expression only in involved areas.

Th1, type 1 Epithelial barrier defect?

Defective epithelial barrier function; activation of mucosal macrophages

102

Over-expression of HLA-B27

Inflammation of stomach as well as small and large intestine; joint inflammation

Th2, type 1;

Facilitated presentation of mucosal antigens to mucosal T cells; CD8+ T cell-mediated inflammation?

205, 206

NF-κB defects; “A20” mouse; Iκβα-deficiency

Inflammation in multiple organs including intestine

Type 1

Hypersensitivity to NF-κB activators; inability to regulate NF-κB response

207, 208

p50 deficiency

Typhlo-colitis; apparently normal T cell development

Type 2?

Defect in NF-κB pathway

209

Th2 TNBS-colitis

TNBS-colitis in Balb/c mice or C57BL/6 mice with IL-12 deficiency

Th2 colitis; Th1 component?; type 1

TNBS-induced, Th2 response to mucosal antigens in mice oriented to Th2 responses

210, 211

Colitis in WiskottAldrich syndrome protein (WASP) deficiency

Mild immunodeficiency; superficial inflammation reminiscent of UC rather than CD

Th2 colitis; Type 2?

Abnormality of regulatory cell development?

212

Abbreviations: TCR, T cell receptor; TNBS, trinitrobenzene sulfonic acid; UC, ulcerative colitis; CD, Crohn’s disease.

The focus of research utilizing this model has been to define the relation of bacterial flora to the induction of disease. A series of studies with this in mind showed that cells from C3H/HeJBir mice manifest increased B cell and T cell reactivity to mucosal antigens (92, 170). However, this reactivity was selective and was more or less limited to antigens associated with several species of facultative anaerobes. This corresponded to the fact that T cells in colitic C3H/HejBir mice displayed a skewed Vβ distribution (as do T cells in other models including the SCID-transfer model and the TCR-α chain–deficiency model) (92). It should be noted, however, that although the number of antigens implicated in the response is a small fraction of the total number of antigens, it is nevertheless a large number; thus, it is highly unlikely that the colitis is due to only a very limited number of antigens. A similar situation probably obtains in humans, in which a skewed expression of Vβ has also been seen.

4 Feb 2002

11:44

Annu. Rev. Immunol. 2002.20:495-549. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

526

AR

AR152-17.tex

STROBER

¥

FUSS

AR152-17.SGM

¥

LaTeX2e(2001/05/10)

P1: GKD

BLUMBERG

Recent cell transfer studies of the C3H/HeJBir model have disclosed that T cell lines driven by antigens in bacterial lysates also transfer colitis to SCID mice (91). This finding is a direct demonstration that antigens in the mucosal microflora can mediate colitis and corroborates data from studies of numerous other models that show this more indirectly. Whereas some of the lines were composed of effector cells secreting IFN-γ , others were slow-growing lines producing IL-10 and were presumed to be regulatory cells. Indeed, these cells inhibited Th1 responses both in vivo and in vitro. Thus, a reiteration of studies performed in the SCID-transfer model showed that these cell lines, when cotransferred with effector cells, could prevent the development of colitis in SCID recipients (93). Interestingly, this preventative effect was reversed by either IL-10 or TGF-β, again raising questions about the relationship of these regulatory cytokines. In any case, the presence of regulatory cells in the colitis of C3H/HeJBir mice suggests that the underlying defect in these mice is a partial block in the regulatory cells’ development and that the inflammation gradually subsides when these cells finally make their appearance. One issue raised by the occurrence of colitis in a substrain of mice that does not respond to LPS is the role of this stimulant in the regulation of mucosal immune responses. It appears paradoxical that a defect in the capacity to respond to a strong stimulator of the IL-12 response would be associated with colitis. This paradox, however, is resolved by the fact that numerous other substances associated with the mucosal microflora have this capacity as well. Perhaps a more cogent and specific role for LPS in colitis relates to the possibility that this stimulant is necessary for the normal induction of regulatory cells. COLITIS IN SAMP1/Yit MICE SAMP1/Yit mice were originally derived from AKR mice by extensive interbreeding, first to achieve accelerated senescence and then to enhance the development of intestinal inflammation (37). This model is important because the inflammation is remarkably similar to human Crohn’s disease in that it is mainly an ileitis rather than a colitis, and at the microscopic level ones sees typical granulomata and other features of Crohn’s inflammation. Once again the disease is driven by antigens in the mucosal microflora and is a Th1 event because it can be transferred to SCID recipients by T cells producing IFN-γ and TNF-α (37, 38). Interestingly, the transferred cells produce a disease similar to that in the donor mice, suggesting that they have a homing pattern governed by their site of origin or are expanded by antigens specifically present in the ileum. The underlying defect in SAMP1/Yit mice is not yet known. However, recent studies showing that epithelial cells in these mice produce increased amounts of chemokines suggest the presence of a type 1 defect (171).

MODELS OF COLITIS DUE TO EXCESSIVE Th2 T CELL RESPONSES Experimental colitis mediated by Th2 T cells forms a separate universe of colitides that, as discussed above, are associated with a form of inflammation that differs from that seen in the more predominant Th1 colitides in that it more closely

4 Feb 2002

11:44

AR

AR152-17.tex

AR152-17.SGM

LaTeX2e(2001/05/10)

P1: GKD

Annu. Rev. Immunol. 2002.20:495-549. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

INSIGHTS INTO IBD

527

resembles ulcerative colitis than Crohn’s disease (23). This has given rise to the idea that ulcerative colitis is in fact a Th2 T cell–mediated disease, but the evidence for this notion is ambiguous at best. Thus, whereas IL-12/IFN-γ production is not increased in ulcerative colitis, neither is IL-4 production, and the inflammation in lesional tissue is not usually characterized by Th2 inflammatory elements such as eosinophils and mast cells. In fact, the only Th2 cytokine reported as increased in ulcerative colitis is IL-5 (39–42), but this increase could be due to the presence of certain types of regulatory cells that produce IL-5 rather than to Th2 effector cells (i.e., Tr1 T cells). At the moment, therefore, it is premature to call ulcerative colitis a Th2 disease, despite its histopathologic relation to Th2 T cell–mediated experimental colitides. In the following discussion we review the most extensively studied Th2 models; pertinent data about several additional Th2 models is provided in Table 4.

Colitis in TCR-α Chain–Deficient Mice The first and perhaps best studied model of murine inflammation owing to a Th2mediated response was initially reported by Mombaerts et al., who noted that chronic colitis develops in gene targeted mice lacking TCR-α chains (51). These authors also found that TCR-β chain–deficient mice develop only very mild colitis, but recently it was shown that this colitis is more marked if CD5 deficiency is also present (172). Finally, they showed that mice with γ δ chain deficiency do not develop colitis. The colitis developing in TCR-α chain–deficient mice is relatively superficial and extends to the submucosa only occasionally. It is characterized by the presence of elongated and distorted crypts as well as by the presence of occasional crypt abscesses, but transmural fissures and granulomata are notably absent. Overall the lesion is different from that seen in Th1 models of colitis and resembles ulcerative colitis rather than Crohn’s disease. This fits with the fact that affected mice frequently develop circulating anti-neutrophil cytoplasmic antibodies (ANCA) and other antibodies found in ulcerative colitis patients (107, 173). Initial studies of the cell populations in TCR-α chain–deficient mice revealed that the main cells were γ δ TCR-bearing T cells, but these were admixed with a small population of alpha-beta+ (dim) TCR-bearing T cells (ββ TCR T cells), which later proved to be the effector cells responsible for the inflammation (49, 51). In subsequent studies involving treatment of TCR-α chain–deficient mice with anticytokine antibodies and cross-breeding of the mice with various cytokinedeficient mice, it was established that IL-4 and not IFN-γ was the effector cytokine, i.e., the colitis was a Th2 colitis (174–176). The ββ TCR T cells established as effector T cells in TCR-α chain colitis recognize antigens via a unique TCR composed of ββ homodimers, and it was therefore not too surprising that they display greatly restricted TCR diversity following the development of colitis (34). Thus, whereas T cells obtained from mice prior to the development of colitis or from mice on a non–disease-producing elemental diet display a wide range of Vβ family usage, those with colitis or on a regular diet

4 Feb 2002

11:44

Annu. Rev. Immunol. 2002.20:495-549. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

528

AR

AR152-17.tex

STROBER

¥

FUSS

AR152-17.SGM

¥

LaTeX2e(2001/05/10)

P1: GKD

BLUMBERG

display skewed Vβ usage marked by Vβ 8.2 predominance (34). In addition, it was found in single-strand conformation polymorphism (SSCP) analysis and CDR3 sequencing studies that T cells from colitic mice display mono- or oligoclonality in all T cell subsets, not just the Vβ 8.2 T cell subset. Finally, it was shown that T cell subsets expressing Vβ 8.2 exhibiting restricted diversity are characterized by a restricted CDR3 length and conservation of a single negatively charged amino acid in the second portion of the CDR3 sequence (33). This type of amino acid sequence is characteristic of clones specific for self-antigens and thus may represent a “germ-line” sequence that is cross-reactive with a variety of normally nonstimulatory environmental (mucosal) antigens, i.e., antigens that do not elicit effector cell responses in the mucosal immune system. This possibility is supported by the fact that cells with the stereotypic TCRs can be expanded by coculture with colonic epithelial cells, which presumably are presenting antigens derived from resident (nonpathogenic) microflora (33). It is also supported by the facts that ββ TCR T cells display vigorous responses to food antigens and that mice with TCR-α chain deficiency exhibit a heightened capacity to provide helper function for B cells that produce antibodies reacting to food antigens (34, 175). Finally, it is supported by the fact that TCR-α chain–deficient mice fed an elemental diet do not develop disease unless they are mono-infected with certain organisms (such as B. vulgatus) that are presumably among the organisms expressing cross-reacting antigens (97). The above data, considered as a whole, lead to the conclusion that whereas ββ TCR T cells may have a restricted ability to respond to antigens in general, they do respond to certain normally harmless antigens that then drives the cells to expand and exert effector cell activity causing disease. This raises the question of why these T cells have this propensity, but other T cells normally populating the mucosal tissues do not. One possibility already suggested by the TCR sequence data is that these cells have escaped negative selection in the thymus (or other selection areas exisiting in the mucosa) and thus represent a cadre of self-reactive cells that cross-react with mucosal antigens. This possibility finds strong support in independent studies showing that ββ TCR T cells have a tendency to escape negative selection in the thymus (177, 178). A second possibility relates to the fact that, as discussed above, mucosal responses are normally regulated by tolerogenic mechanisms, including the development of αβ TCR T cells that produce suppressive cytokines. Thus, it is reasonable to suggest that the abnormal reactivity of ββ TCR T cells to certain mucosal antigens is due to the fact that regulatory T cells cannot develop within the ββ TCR T cell population (nor in the accompanying γ δ TCR T cell population). A second, interrelated question concerns the reason why ββ TCR T cell stimulation leads to Th2 responses and not Th1 responses. One possibility alluded to in the general discussion of models is that the course of T cell differentiation depends largely on the nature of the antigen-presenting cell and the cytokine environment of the APC-T cell interaction. In this context it is already known that certain dendritic cells present in the Peyer’s patches (CD11c+ dendritic cells) preferentially

4 Feb 2002

11:44

AR

AR152-17.tex

AR152-17.SGM

LaTeX2e(2001/05/10)

P1: GKD

Annu. Rev. Immunol. 2002.20:495-549. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

INSIGHTS INTO IBD

529

secrete IL-10 rather than IL-12 and induce Th2 T cells rather than Th1 T cells during antigen presentation (70, 71). Thus, it is possible that the types of antigens that stimulate ββ TCR T cells are generally the types of antigens that are taken up by dendritic cells in Peyer’s patches that lead to Th2 T cell differentiation. Another, non–mutually exclusive, possibility is that ββ TCR T cells manifesting a Th2 phenotype have better survivability than their Th1 counterparts, again because of the cytokine milieu in which they develop. This possibility derives from the observation that stimulation of T cells from TCR-α chain–deficient mice with epithelial cells both under Th1 and Th2 conditions leads to poorer survival in the former instance than in the latter (33). Moreover, the surviving Th2 T cells display evidence of oligoclonality and can transfer disease, whereas the Th1 T cells do not. These data suggest not only that antigens stimulating ββ TCR T cells only do so under a Th2 condition but also that such stimulation under a Th1 condition leads to a different pattern of clonal stimulation and T cells that are subject to apoptosis. A final point to emerge from the study of TCR-α chain–deficient mice relates to the role of B cells in the pathogenesis of this model of inflammation and, by extension, in ulcerative colitis. In particular, it was found that double mutant TCR-α chain–deficient µIg-deficient mice somewhat paradoxically develop more severe colitis than single mutant TCR-α chain–deficient mice (63). Furthermore, transfer of mesenteric lymph nodes (MLN) cells from the double mutant to Rag-2– deficient mice produced colitis in the latter, which was abolished by the cotransfer of B cells and the coadministration of purified Ig or monoclonal antibodies reactive with colonic epithelial cells from the TCR-α chain–deficient mice. This decreased disease with B cells or B cell products (autoantibodies) was associated with decreased numbers of apoptotic cells in the epithelium and lamina propria and was attributed to decreased clearance of these cells mediated by the autoantibodies (63). Another explanation, however, is that the B cells produce regulatory cytokines that suppress disease (and also affect apoptosis). Indeed, preliminary studies suggest that B cells express high levels of CD1d and secrete IL-10 (A. Mizoguchi, R.J. Blumberg & A. Bhan, personal communication). In any case, these studies show that B cells or the autoantibodies they produce do not play a pathogenic role in TCR-α chain–deficient mice and may actually ameliorate disease. In addition, this suggests that autoantibodies in ulcerative colitis are likewise nonpathogenic.

Oxazalone Colitis Whereas administration of TNBS to SJL/J mice leads to colitis driven by polarized Th1 T cell responses, administration of another haptenating agent, oxazalone, leads to a colitis caused by a polarized Th2 T cell response. Oxazalone colitis, however, is a considerably different disease than its TNBS counterpart (31). First, when administered intrarectally without prior sensitization, it develops more quickly and resolves more quickly than TNBS colitis, usually within 4–5 days; in addition, it

4 Feb 2002

11:44

Annu. Rev. Immunol. 2002.20:495-549. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

530

AR

AR152-17.tex

STROBER

¥

FUSS

AR152-17.SGM

¥

LaTeX2e(2001/05/10)

P1: GKD

BLUMBERG

produces a more superficial inflammation that affects the distal half of the colon rather than the whole colon. Finally, rather than producing an intense infiltrative inflammation that obliterates villous architecture, it produces inflammation that generally maintains villous architecture but is associated with bowel wall edema and luminal exudates. Overall, the lesion is more reminiscent of ulcerative colitis than Crohn’s disease, but the examination of a more chronic oxazalone colitis would be necessary to verify this view. If mice are presensitized with subcutaneous oxazalone, a more chronic lesion ensues that lasts on the order of 1–2 weeks; this lesion retains the characteristics described above for the more acute lesion and lends credence to the idea that oxazalone colitis is indeed an ulcerative-colitis-like inflammation (F. Scheiffele, I. Fuss, W. Strober, unpublished observations). The cytokine response of oxazalone colitis is also very different from that in TNBS colitis (31). It is dominated by a high IL-4 and IL-5 response, but a normal or reduced IFN-γ response. This Th2 response is in fact the cause of the inflammation, as shown by the fact that anti–IL-4 administration abolishes disease, whereas an anti–IL-12 administration exacerbates the disease and causes a pancolitis. Another notable feature of the cytokine response in oxazalone colitis is a marked TGF-β response that is higher in the proximal colon than in the distal colon. In fact, the high TGF-β response may be responsible for the short duration of disease as well as its limitation to the distal colon. This is suggested by the fact that, as mentioned above, TGF-β production is higher in the proximal colon, as well as the fact that anti–TGF-β treatment leads to severe pancolitis. The reason lamina propria cells in SJL/J mice respond to oxazalone with a Th2 response rather than a Th1 response is unclear. One cannot invoke the idea that the presence of T cells with abnormal TCRs that only respond to antigens that induce Th2 differentiation because there is nothing to show that the T cell profiles of the responding mice are abnormal. A more likely possibility arises from emerging evidence that during the induction of oxazalone colitis, oxazalone is presented to T cells by APCs in the context of an atypical MHC class I molecule, CD1d, and that the interacting T cell is an NK T cell that is the effector cell causing the colitis (F. Scheiffele, I. Fuss, W. Strober, unpublished observations). Thus, one might postulate that this somewhat unique interaction preferentially results in Th2 T cell differentiation. Some evidence in support of this possibility is inherent in older studies showing that NK T cells have a propensity to produce IL-4, as well as newer studies showing that mice with a targeted deletion of the chemokine receptor, CCR5, when challenged with dextran sulfate sodium (DSS) to produce DSS colitis develop lesions containing T cells producing IL-4 (179). Whereas the reason mice with this deletion manifest this kind of response is not really known, one might postulate that the absence of CCR5 leads to decreased Th1 responses and thus the preferential expansion of NK T cells that inherently produces Th2 cytokines. It should be noted that in CCR5-deficient mice, DSS colitis is less severe than in normal mice, indicating that the NK T cells developing in this situation appear to be regulatory cells. This is in contrast to the situation in oxazalone colitis where, as mentioned, the NK T cells are effector cells. Whether NK T cells act as

4 Feb 2002

11:44

AR

AR152-17.tex

AR152-17.SGM

LaTeX2e(2001/05/10)

P1: GKD

INSIGHTS INTO IBD

531

Annu. Rev. Immunol. 2002.20:495-549. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

regulatory cells in the context of DSS colitis because they produce IL-4 or because of other factors remains to be determined. Aside from the nature and origin of the effector cell causing disease in oxazalone colitis is the question of the downstream inflammatory cytokines causing disease. One possibility that requires further study is that oxazalone colitis leads to production of IL-4, which in turn leads to the secretion of other cytokines, such as IL-9 and IL-13.

MODELS OF COLITIS RELATED TO DEFECTS IN EPITHELIAL CELL BARRIER FUNCTION A number of diverse models of colitis have been discovered that are due to defects in epithelial cell barrier function. It should be emphasized, however, that such defects in the present context are broadly defined to include both barrier function relating to permeability to macromolecules and barrier function involving processes that enable the intestinal epithelial cell to secrete immune mediators. The latter type of defect could take the form of inadequate secretion of mediators that thereby increases the exposure of the mucosal system to antigens in the mucosal microflora or to excessive secretion of mediators and the initiation of inflammation by the stimulation of “professional” cellular secretors of inflammatory cytokines (macrophages). In addition to the models described below, two models already discussed (colitis associated with IL-2 deficiency and with IL-10 deficiency) have been shown to have abnormal epithelial cell barrier function. In these cases it is likely that the latter is secondary to a more primary abnormality as it is known that both Th1 and Th2 cytokines can influence barrier function in various ways.

Colitis Associated with Dominant-Negative N-Cadherin Expression Cadherins are transmembrane glycoproteins that mediate adherence between many cell types including intestinal epithelial cells. On the cytoplasmic side they bind to the cytoskeleton via interactions with β-catenin and on the cell surface, and they enter homophilic interactions with cadherins on neighboring cells (180). Recently, Hermiston & Gordon created a model with disrupted epithelial cell cadherin function by expressing a dominant-negative N-cadherin in epithelial cells that interfere with normal expression of E-cadherin (64). In particular, they inserted embryonic stem cells with an N-cadherin gene lacking an extracellular domain under the control of a small intestinal epithelial cell promotor (the fatty acid binding protein promotor) into blastocysts to obtain chimeric mice that displayed patches of epithelial cells with poor cell-cell adhesion. The chimeric mice developed transmural cellular infiltration, cell crypt abscesses, goblet cell depletion, and both apthous and linear ulcers in lamina propria areas subjacent to the epithelial patches containing

4 Feb 2002

11:44

Annu. Rev. Immunol. 2002.20:495-549. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

532

AR

AR152-17.tex

STROBER

¥

FUSS

AR152-17.SGM

¥

LaTeX2e(2001/05/10)

P1: GKD

BLUMBERG

cells with defective adherence but not in areas subjacent to epithelial patches with normally adherent epithelial cells. The most reasonable explanation for this pattern of inflammation is that in areas of the mucosa where there is breakdown of the epithelial barrier (owing in this case to defective cell-cell adhesion) there is excessive exposure of mucosal lymphoid elements (normal mucosal microflora), which subsequently leads to a nonhomeostatic immune response and mucosal inflammation. This picture is thus not unlike that in TNBS colitis, in which the introduction of TNBS in the presence of the substance (ethanol) that disrupts the mucosal barrier leads to an unbalanced immune response and subsequent inflammation. As to the question of why such exposure leads to inflammation, it can be postulated that any exposure of the mucosal immune system to antigens in a manner that bypasses the Peyer’s patches leads to an inadequate regulatory T cell response because such cells preferentially develop in the organized lymphoid tissue of the mucosa. Finally, because the N-cadherin dominant-negative model of inflammation occurs in the vicinity of porous epithelial cells despite the fact that the microbial microflora are identical in both nonporous and porous areas, it is an exquisite demonstration of the fact that antigens in normal mucosal microflora are sufficient for the induction of responses that lead to disease.

Colitis in mdr1a-Deficient Mice A second and equally interesting model of colitis related to barrier function is a mouse model characterized by deficiency in the mdr1a gene (65). This gene is one of several “multiple drug-resistant” (mdr) genes expressed in many cells (including epithelial cells) that belong to a family of transporter proteins that pump small amphiphilic and hydrophobic molecules out of the cell and thus confer drug resistance (181, 182). This model was created because the gene encoding the mdr1a transporter is present in a region of the human genome that is thought to harbor a disease gene that leads to inflammatory bowel disease (183, 184). Bone marrow transfer studies involving wild-type donor cells into mdr1a-deficient recipients demonstrated that the colitis develops in mdr1a-deficient mice because of the deficiency of mdr1a in epithelial cells rather than in lymphoid or myeloid cells (65). Thus, the model allows one to focus on the role of epithelial cells in mucosal inflammation. Colitis developing in mdr1a-deficient mice is a spontaneous colitis consisting of a transmural T cell and B cell infiltration that is similar to that found in Crohn’s disease (despite the fact that it has been called a model of ulcerative colitis). It is important to mention, however, that the epithelial cells in mdr1a-deficient mice are arrayed in long, dysregulated crypts that are associated with crypt abscesses and surface ulcerations (65). The Crohn’s disease–like picture was borne out by the presence of an mRNA cytokine profile indicative of a Th1-mediated inflammation (J. Viney, personal communication). Recently, it has been shown that mucosal organisms can profoundly influence epithelial cell function with respect to cytokine/chemokine secretion. This influence can be negative in that epithelial cell interactions with certain nonpathologic

4 Feb 2002

11:44

AR

AR152-17.tex

AR152-17.SGM

LaTeX2e(2001/05/10)

P1: GKD

Annu. Rev. Immunol. 2002.20:495-549. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

INSIGHTS INTO IBD

533

organisms leads to a block in ubiquitination of Iκβα (necessary for Iκβα degradation) and thus a subsequent block in NF-κB translocation to the nucleus (185). This influence can also be positive because bacterial flagellin, signaling through TLR-5, leads to NF-κB expression (186). If these events are influenced or caused by bacterial products entering the cell, we can see how epithelial cell function can be negatively or positively impacted by a defect in a transporter mechanism. Furthermore, we can understand how such negative or positive perturbation could lead to increased epithelial cell production of chemokines and cytokines that lead to the influx of inflammatory elements into the epithelial layer. Thus, it seems possible that the mdr1a-deficient mouse does not develop inflammation because of increased epithelial layer permeability per se but because of increased bacterially induced activation of epithelial cells.

Dextran Sulfate Colitis Yet another model of colitis that is at least partially related to a change in epithelial cell barrier function is the colitis induced by the physical agent, dextran sulfate sodium. This is a relatively old model that has been frequently used to study the efficacy of potential therapeutic agents because of its ease to induce via administration of DSS in drinking water and because DSS induces a consistent level of colitis with a defined onset (18–22). As mentioned in the Introduction, DSS colitis can be induced in Rag-2–deficient or thymectomized mice (22). This argues that the mechanisms of inflammation in this form of colitis are, at least initially, the activation of nonlymphoid cells such as macrophages and the release of pro-inflammatory cytokines (187, 188). Changes in epithelial barrier function as measured by permeability of the intestinal wall to Evan’s blue can be found early (several days before the onset of frank inflammation) (19) and thus may set the stage for macrophage activation. The relation of DSS colitis to epithelial barrier function is further suggested by the fact that administration of DSS to mice with deficiency of intestinal trefoil factor, a factor important to maintenance and repair of the epithelial layer, leads to a far more severe colitis than observed in normal mice (189). However, this may be a result of the fact that epithelial cell layer integrity plays a role in the initiation of DSS colitis as indicated above, or because reestablishment of such integrity is a condition of colitis resolution. In the acute stages of DSS colitis the (secondary?) T cell response consists of a polarized Th1 response, but in later and more chronic phases of the inflammation, a mixed Th1/Th2 response occurs (188). In either case, DSS elicits the secretion of large amounts of TNF-α and IL-6, which are mainly responsible for the tissue damage in the disease. Whereas antigens in the mucosal microflora probably play a role in the production of DSS colitis, it has recently been shown that they also play a role in the suppression (and resolution) of the colitis. This is shown by the fact that mice administered α-galacocylceramide (a glycolipid antigen that activates NK T cells when presented to them in the context of CD1d, a nonclassical MHC class I antigen-presenting molecule expressed on epithelial cells and other APCs)

4 Feb 2002

11:44

Annu. Rev. Immunol. 2002.20:495-549. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

534

AR

AR152-17.tex

STROBER

¥

FUSS

AR152-17.SGM

¥

LaTeX2e(2001/05/10)

P1: GKD

BLUMBERG

manifest decreased DSS colitis as compared with mice administered a control glycolipid without these properties. This improvement is not seen in mice deficient in CD1d or in Rag−/− negative mice deficient in T cells (and NK T cells) (61). Finally, by confocal microscopy the administered α-galactocylceramide could be localized to the epithelium. These studies reveal that NK T cells stimulated by glycolipid antigen play a protective role in DSS colitis that is not unlike their role in the SCID colitis model mentioned above. The mechanism of this role is not yet defined, but given the localization of the stimulating antigen to the epithelium, the NK T cells may be secreting substances that reestablish epithelial integrity or regulate other aspects of the mucosal environment that normally drive the colitis in this model. As discussed in relation to NK T cells found in CCR5-deficient mice, this may involve the secretion of Th2 cytokines.

TREATMENT OF MODELS OF MUCOSAL INFLAMMATION The fact that models of mucosal inflammation, whatever their underlying cause, resolve themselves into either Th1 or Th2 T cell–mediated inflammation has led to the recognition that models of vastly different etiologies can be treated with agents that block these final common pathways at any of a variety of points. As shown in Figure 2, “points of attack” in the Th1 pathway can readily be identified and can be used to block the pathway in both models of inflammation and human Crohn’s disease. A similar diagram can be drawn with respect to the Th2 pathway, which can be applied to Th2 models of inflammation and perhaps human ulcerative colitis. Only the broad outline of such treatment approaches can be discussed in this review. With respect to Th1-mediated inflammation, the use of agents that block IL-12 secretion or IL-12 activity provides the most direct approach because, as we have seen, depriving Th1 cells of IL-12 leads to their apoptosis (32, 123). It should be noted that not only anti-IL-12, but also other agents that downregulate IL-12 secretion are possible therapeutic agents in this context (190, 191). A related kind of therapy involves the use of anti–TNF-α antibody and soluble TNF-R agents that, as discussed above, block the Th1 response at both the inductive and the effector phases of the response. This approach has already proven useful in the treatment of human Crohn’s disease (192). A parallel approach to Th2-mediated inflammation is more problematic in that although anti–IL-4 may be an effective treatment of Th2 models, it does not apply to human ulcerative colitis because this disease has not been shown to be caused by IL-4 dysregulation. A promising alternative approach is to target IL-6R with the use of an anti–IL-6 receptor antibody (193, 194). Such therapy blocks IL-6 “transsignaling” and leads to the apoptosis of both Th1 and Th2 T cells. Thus, it is theoretically applicable to both Th1- and Th2-mediated inflammation. It is also possible that models of inflammation can be treated with regulatory (suppressive) cytokines such as IL-10 or TGF-β. In studies conducted so far IL-10 has been applied with mixed success and, likewise, has been only marginally effective in human IBD (195, 196). The problem may be one of cytokine localization

4 Feb 2002

11:44

AR

AR152-17.tex

AR152-17.SGM

LaTeX2e(2001/05/10)

P1: GKD

Annu. Rev. Immunol. 2002.20:495-549. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

INSIGHTS INTO IBD

535

Figure 2 As shown in this diagram of Th1 T cell–mediated mucosal inflammation, the Th1 pathway can be “attacked” (i.e., inhibited or disrupted) at many different points, each representing a potential means of therapeutic intervention. These points are defined as follows: 1) inhibitors of IL-12/IL-18 (i.e., anti–IL-12, rCT-B, or β-agonists); 2) inhibitors of DC-T cell interaction (i.e., anti–CD40L or anti–CD134); 3) inhibitors of TNF-α (i.e., anti– TNF-α and TNF-αR); 4) IL-10 and TGF-β (i.e., TGF-β or IL-10 plasmids or administration of Th3 or Tr1 cells); 5) inhibition of IL-6 trans-signaling, anti–IL-6R; 6) NF-κB inhibitors; 7) inhibitors of homing or adhesion (i.e., anti-α4β7, anti-αEβ7, anti-CD44v7, or anti-sense oligos to ICAM1); 8) downstream inhibitors of TNF-α (i.e., phosphodiasterase inhibitor 4, pentoxyphylline, thalidomide, or metalloproteinase inhibitors).

because there is one report that IL-10 delivered by a Lactococcus lactis organisms was effective in two forms of experimental colitis (197). The use of TGF-β has been explored in studies in which plasmids encoding TGF-β are administered intranasally (29). As mentioned above, this leads to cells producing TGF-β that migrate to mucosal tissues that are capable of reversing established TNBS colitis. A possible objection to this approach is that TGF-β can induce fibrosis; however, the plasmid also induces IL-10 secretion, which appears to suppress TGF-β–induced fibrosis (198). Other approaches to the treatment of models of mucosal inflammation include the use of agents that target homing and localization of inflammatory cells. This

4 Feb 2002

11:44

536

AR

AR152-17.tex

STROBER

¥

FUSS

AR152-17.SGM

¥

LaTeX2e(2001/05/10)

P1: GKD

BLUMBERG

includes antibodies to integrins (anti–MadCAM-1, anti-α Eβ 7 and anti-CD44v7 as well as anti-sense oligonucleotides that interfere with integrin synthesis (199– 201). Finally, attempts to control the mucosal inflammation by the use of agents that block the NF-κB pathway have been tested with some success in murine models (202–203). The question is whether such therapy will cause unacceptable toxicity when applied to humans.

Annu. Rev. Immunol. 2002.20:495-549. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

CLOSING CONSIDERATIONS In this review of models of mucosal inflammation we have sought to emphasize recurrent characteristics of the models that allow them to be understood within a more or less consistent framework. This is perhaps best encapsulated by the fact that the models are invariably associated with one or another genetically determined or induced immune imbalance that ultimately expresses itself as a type 1 defect owing to excessive effector cell response or a type 2 defect owing to an inadequate regulatory cell response. Furthermore, in both the type 1 and type 2 defects, the response takes the form of either a Th1 or Th2 T cell-mediated inflammation that is driven not by antigens associated with exogenous pathogenic organisms but by antigens associated with the normal mucosal microflora. Given the fact that such antigens are the equivalent of self-antigens, the models may thus be visualized as a special type of autoimmunity that takes on a somewhat unique form because it involves effector and regulatory cell mechanisms that are characteristic of mucosal responses. The impact of the knowledge gained from the study of models of inflammation on the understanding of human IBDs is difficult to exaggerate. Thus, it is fair to say that the framework used to visualize the pathogenesis of these diseases is currently derived largely from the murine models and, in turn, new patient-oriented research is mainly motivated by one or another aspect of the models. This includes research on new treatments of the disease that are either suggested by the models or are tested in the models. Looking ahead to the emerging area of genetic research in IBDs, the models will be an essential tool in the identification of genes that determine susceptibility and resistance to these diseases and thus the genes that will enable their genetic manipulation. Visit the Annual Reviews home page at www.annualreviews.org

LITERATURE CITED 1. Strober W. 1985. Animal models of inflammatory bowel disease—an overview. Dig. Dis. Sci. 30:(Suppl.):3–8 2. Kim HS, Berstad A. 1992. Experimental colitis in animal models. Scand. J. Gastroenterol. 27:529–37

3. Kirsner JB. 1961. Experimental “colitis” with particular reference to hypersensitivity reactions in the colon. Gastroenterology 40:307–12 4. Kraft SC, Fitch FW, Kirsner JB. 1963. Histologic and immunohistochemical

4 Feb 2002

11:44

AR

AR152-17.tex

AR152-17.SGM

LaTeX2e(2001/05/10)

P1: GKD

INSIGHTS INTO IBD

5.

Annu. Rev. Immunol. 2002.20:495-549. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

6.

7.

8.

9.

10.

11.

12.

13.

14.

15.

features of Auer “colitis” in rabbits. Am. J. Pathol. 43:913–23 Mee AS, McLaughlin JE, Hodgson HJ, Jewell DP. 1979. Chronic immune colitis in rabbits. Gut 20:1–5 Halpern B, Zweibaum A, Oriol Palou R, Morard JC. 1967. Experimental immune ulcerative colitis. In Immunopathology, International Symposium, pp. 161– 78. Basel/Stuttgart: Schwabe and Co. Rabin BS, Rogers SJ. 1978. A cellmediated immune model of inflammatory bowel disease in the rabbit. Gastroenterology 75:29–33 Glick ME, Falchuk ZM. 1981. Dinitrochlorbenzene-induced colitis in the guinea-pig: studies of colonic lamina propria lymphocytes. Gut 22:120–25 MacPherson BR, Pfeiffer CJ. 1978. Experimental production of diffuse colitis in rats. Digestion 17:135–50 Sharon P, Stenson WF. 1985. Metabolism of arachidonic acid in acetic acid colitis in rats: similarity to human inflammatory bowel disease. Gastroenterology 88:55– 63 Fretland DJ, Widomski DL, Levin S, Gaginella TS. 1990. Colonic inflammation in the rabbit-induced by phorbol-12myristate-13-acetate. Inflammation 14: 143–50 Marasco WA, Phan SH, Krutzch H, et al. 1984. Purification and identification of formyl-methionyl-leucyl-phenylanine as the major peptide neutrophil chemotactic factor produced by Escherichia coli. J. Biol. Chem. 259:5430– 39 Williams LT, Synderman R, Pike MC, Lefkowitz RJ. 1977. Specific receptor site for chemotactic peptides on human polymorphonuclear leukocytes. Proc. Natl. Acad. Sci. USA 74:1204–8 Painter GR, Sklar LA, Jesatis AJ, Schmitt M, Cochrane CG. 1984. Activation of neutrophils by N-formyl chemotactic peptides. Fed. Proc. 43:2737–42 Marcus R, Watt J. 1969. Seaweeds and

16.

17.

18.

19.

20.

21.

22.

23.

24. 25.

26.

537

ulcerative colitis in laboratory animals. Lancet 2:489–90 Abraham R, Fabian RJ, Goldberg L, Coulston F. 1974. Role of lysosomes in carageenan-induced cecal ulceration. Gastroenterology :1169–81 Watt J, Marcus R. 1972. Ulceration of the colon in rabbits fed sulfated amylopectin. J. Pharm. Pharmacol. 24:68–69 Ohkusa T. 1985. Production of experimental ulcerative colitis in hamsters by dextran sulfate sodium and change in intestinal microflora. Jpn. J. Gastroenterol. 82:1327–36 Kitajima S, Takoma S, Morimoto M. 1999. Changes in colonic mucosal permeability mouse colitis induced with dextran sulfate sodium. Exp. Animal 48:137–43 Onderdonk AB, Hermos JA, Dzink JL, Bartlett JG. 1978. Protective effect of metronidazole in experimental ulcerative colitis. Gastroenterology 74:521–26 Rath HC, Schultz M, Freitag R, et al. 2001. Different subsets of enteric bacteria induce and perpetuate experimental colitis in rats and mice. Infect. Immun. 69:2277–85 Axelsson LG, Landstrom E, Goldschmitt TJ, Gronberg A, Bylund-Fellenius AC. 1996. Dextran sulfate sodium (DSS) induced experimental colitis in immunodeficient mice: effects in CD4+-cell depleted, athymic and NK-cell depleted SCID mice. Inflamm. Res. 45:181–91 Strober W, Fuss IJ, Ehrhardt RO, Neurath M, Boirivant M, Ludviksson BR. 1998. Mucosal immunoregulation and inflammatory bowel disease: new insights from murine models of inflammation. Scand. J. Immunol. 48:453–58 Strober W, Kelsall B, Marth T. 1998. Oral tolerance. J. Clin. Immunol. 18:1–30 Ludviksson BR, Seegers D, Resnick AS, Strober W. 2000. The effect of TGFbeta1 on immune responses of na¨ıve versus memory CD4+ Th1/Th2 T cells. Eur. J. Immunol. 30:2101–11 Seder RA, Marth T, Sieve MC, Letterio

4 Feb 2002

11:44

538

Annu. Rev. Immunol. 2002.20:495-549. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

27.

28.

29.

30.

31.

32.

33.

AR

AR152-17.tex

STROBER

¥

FUSS

AR152-17.SGM

¥

LaTeX2e(2001/05/10)

P1: GKD

BLUMBERG

JJ, Roberts AB, Kelsall B. 1998. Factors involved in the differentiation of TGFbeta-producing cells from na¨ıve CD4+ T cells: IL-4 and IFN-γ have opposing effects, while TGF-beta positively regulates its own production. J. Immunol. 160:5719–28 Marth T, Strober W, Seder RA, Kelsall BL. 1997. Regulation of transforming growth factor-beta production by interleukin-12. Eur. J. Immunol. 27:1213–20 Pardoux C, Ma X, Gobert S, Pellegrini S, Mayeux P, Gay F, Trinchieri G, Chouaib S. 1999. Downregulation of interleukin-12 (IL-12) responsiveness in human T cells by transforming growth factor-beta: relationship with IL-12 signaling. Blood 93:1448–55 Kitani A, Fuss IJ, Nakamura K, Schwartz OM, Usui T, Strober W. 2000. Treatment of experimental (trinitrobenze sulfonic acid) colitis by intranasal administration of transforming growth factor (TGF)beta1 plasmid: TGF-beta1-mediated suppression of T helper cell type 1 response occurs by interleukin (IL)-10 induction and IL-12 receptor beta2 chain downregulation. J. Exp. Med. 192:41–52 Xu H, Zhang GX, Wysocka M, Wu CY, Trinchieri G, Rostami A. 2000. The suppressive effect of TGF-beta on IL-12mediated immune modulation specific to a peptide Ac1-11 of myelin basic protein (MBP): a mechanism involved in inhibition of both IL-12 receptor beta1 and beta2. J. Neuroimmunol. 108:53–63 Boirivant M, Fuss IJ, Chu A, Strober W. 1998. Oxazalone colitis: a murine model of T helper cell type 2 colitis treatable with antibodies to interleukin 4. J. Exp. Med. 188:1929–39 Neurath MF, Fuss I, Kelsall BL, St¨uber E, Strober W. 1995. Antibodies to interleukin 12 abrogate established experimental colitis in mice. J. Exp. Med. 182:1281–90 Mizoguchi A, Mizoguchi E, Saubermann LJ, Higaki K, Blumberg RS, Bhan AK.

34.

35.

36.

37.

38.

39.

40.

2000. Limited CD4 T-cell diversity associated with colitis in T-cell receptor alpha mutant mice requires a T helper 2 environment. Gastroenterology 119:983–95 Takahashi I, Iijima H, Katashima R, Itakura M, Kiyono H. 1999. Clonal expansion of CD4+ TCRbeta beta+ T cells in TCR alpha-chain-deficient mice by gutderived antigens. J. Immunol. 162:1843– 50 Spencer DM, Banerjee S, Veldmann GM, Levine AD. 1999. Anti-IL-12 therapy in IL-10 deficient mice prevents the onset of colitis and reverses early diseases, but is ineffective in late disease. Gastroenterology 116:G3298 (Abstract) Kontoyiannis D, Pasparakis M, Pizarro T, Cominelli F, Kollias G. 1999. Impaired on/off regulation of TNF biosynthesis in mice lacking TNF AU-rich elements: implications for joint and gut-associated immunopathologies. Immunity 10:387–98 Matsumoto S, et al. 1998. Inflammatory bowel disease-like enteritis and caecitis in a senescence accelerated mouse P1/Yit strain. Gut 43:71–78 Kosiewicz MM, Nast CC, Krishnan A, Rivera-Nieves J, Moskaluk CA, Matsumoto S, Kozaiwa K, Cominelli F. 2001. Th1-type responses mediate spontaneous ileitis in a novel murine model of Crohn’s disease. J. Clin. Invest. 107:695–702 Fuss IJ, Neurath MF, Boirivant M, Klein JS, de la Motte C, Strong SA, Fiocchi C, Strober W. 1996. Disparate CD4+ lamina propria (LP) lymphokine secretion profiles in inflammatory bowel disease. Crohn’s disease LP cells manifest increased secretion of IFN-γ , whereas ulcerative colitis LP cells manifest increased secretion of IL-5. J. Immunol. 157:1261–70 Mullin GE, Maycon ZR, Braun-Elwert L, Cerchia R, James SP, Katz S, Weissman GW, McKinley MJ, Fisher SE. 1996. Inflammatory bowel disease mucosal biopsies have specialized lymphokine mRNA profiles. Inflamm. Bowel Dis. 2:16–26

4 Feb 2002

11:44

AR

AR152-17.tex

AR152-17.SGM

LaTeX2e(2001/05/10)

P1: GKD

Annu. Rev. Immunol. 2002.20:495-549. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

INSIGHTS INTO IBD 41. Monteleone G, Biancone L, Marasco R, Marrone G, Marasco O, Luzza F, Pallone F. 1997. Interleukin-12 is expressed and actively released by Crohn’s disease intestinal lamina propria mononuclear cells. Gastroenterology 112:1169–78 42. Parronchi P, Romagni P, Annunziato F, Sampognaro S, Becchio A, Giannarini L, Maggi E, Pupilli C, Tonelli F, Romagni S. 1997. Type 1 T-helper cell predominance and IL-12 expression in the gut of patients with Crohn’s disease. Am. J. Pathol. 150:823–32 43. Malmstrom V, Shipton D, Singh B, AlShamkhani A, Puklavec MJ, Barclay AN, Powrie F. 2001. CD134L expression on dendritic cells in the mesenteric lymph nodes drive colitis in T cell-restored SCID mice. J. Immunol. 166:6972–81 44. Takeda K, Clausen BE, Kaisho T, Tsujimura T, Terada N, Forster I, Akira S. 1999. Enhanced Th1 activity and development of chronic enterocolitis in mice devoid of STAT3 in macrophages and neutrophils. Immunity 10:39–49 45. Okamoto S, Watanabe M, Yamazaki M, Yajima T, Hayashi T, Ishii H, Mukai M, Yamada T, Noriaki W, Jameson BA, Hibi T. 1999. A synthetic mimetic of CD4 is able to suppress disease in rodent model of immune cells. Eur. J. Immunol. 29:355– 66 46. Simpson SJ, Mizoguchi E, Allen D, Bhan AK, Terhorst C. 1995. Evidence that CD4+ but not CD8+ T cells are responsible for murine interleukin-2-deficient colitis. Eur. J. Immunol. 25:2618–25 47. Simpson SJ, de Jon YP, Shah SA, Comiskey M, Wang B, Spielman JA, Podack ER, Mizoguchi E, Bhan AK, Terhorst C. 1998. Consequences of Fas-ligand and perforin expression by colon T cells in a mouse model of inflammatory bowel disease. Gastroenterology 115:849–55 48. Strater J, Wellisch I, Riedl S, Walczak H, Koretz K, Tandara A, Krammer PH, Moller P. 1997. CD95 (APO-1/Fas)mediated apoptosis in colon epithelial

49.

50.

51.

52.

53.

54.

55.

56.

57.

539

cells: a possible role in ulcerative colitis. Gastroenterology 113:160–67 Mizoguchi E, Mizoguchi A, Bhan AK. 1997. Role of cytokines in the early stages of chronic colitis in TCRα-mutant mice. Lab. Invest. 76:385–97 Miyashita-Kawaguchi M, Shin-ichiro S, Korosu H, Kato-Nagoka N, Matsuoko Y, Ohwukim M, et al. 2001. An accessory role of TCR γ δ cells in the exacerbation of inflammatory bowel disease in TCRα mutant mice. Eur. J. Immunol. 31:980–88 Mombaerts P, Mizoguchi E, Grusby MJ, Glimcher L, Bhan AK, Tonegawa S. 1993. Spontaneous development of inflammatory bowel disease in T cell receptor mutant mice. Cell 75:275–82 Hoffmann JC, Peters K, Henschke S, Hermann B, Pfister K, Westermann J, Zeitz M. 2001. Role of T lymphocytes in rat 2, 4, 6-trinitrobenzene sulphonic acid induced colitis: increased mortality after γ δ T cell depletion and no effect of αβ T cell depletion. Gut 48:489–95 Boismenu R, Havren WL. 1994. Modulation of epithelial cell growth by intraepithelial γ δ T cells. Science 266:1253– 55 Seder RA, Marth T, Sieve MC, Strober W, Letterio JJ, Roberts AB, Kelsall B. 1998. Factors involved in the differentiation of TGF-beta-producing cells from na¨ıve CD4+ T cells: IL-4 and IFN-γ have opposing effects, while TGF-beta positively regulates its own production. J. Immunol. 160:5719–28 Groux H, O’Garra A, Bigler M, Rouleau M, Antonenko S, de Vries JE, Roncarolo MG. 1997. A CD4+ T-cell subset inhibits antigen-specific-T-cell responses and prevents colitis. Nature 389:737–42 Levings MK, Sangregorio R, Galbiati F, Squadrone S, de Waal Maleft R, Roncarolo MG. 2001. IFN-alpha and IL-10 induce the differentiation of human type 1 T regulatory cells. J. Immunol. 166:5530– 39 Shevach EM, Thornton AM. 1998. CD4+

4 Feb 2002

11:44

540

Annu. Rev. Immunol. 2002.20:495-549. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

58.

59.

60.

61.

62.

63.

64.

65.

66.

AR

AR152-17.tex

STROBER

¥

FUSS

AR152-17.SGM

¥

LaTeX2e(2001/05/10)

P1: GKD

BLUMBERG

CD25+ immunoregulatory T cells suppress polyclonal T cell activation in vitro by inhibiting interleukin 2 production. J. Exp. Med. 188:287–96 Takashi T, Kuniyasu Y, Toda M, Sakaguchi N, Itoh M, Iwata M, Shimuzu J, Sakaguchi S. 1998. Immunologic selftolerance maintained by CD25+ CD4+ naturally anergic and suppressive T cells: induction of autoimmune disease by breaking their anergic/suppressive state. Int. Immunol. 10:1969–80 Nakamura K, Kitani A, Strober W. 2001. Cell contact-dependent immunosuppression by CD4+/CD25+ regulatory T cells is mediated by cell surface-bound transforming growth factor-β. J. Exp. Med. 194:629–44 Fort MM, Leach MW, Rennick DM. 1998. A role for NK cells as regulators of CD4+ T cells in a transfer model of colitis. J. Immunol. 161:3256–61 Saubermann LJ, Beck P, De Jong YP, Pitmann RS, Ryan MS, Kim HS, et al. 2000. Activation of natural killer T cells by alpha-galactosylceramide in the presence of CD1d provides protection against colitis in mice. Gastroenterology 119:119–28 Blumberg RS, Gerdes D, Chott A, Porcelli SA, Balk SP. 1995. Structure and function of the CD1 family of MHClike cell surface proteins. Immunol. Rev. 147:5–29 Mizoguchi A, Mizoguchi E, Smith RN, Preffer FI, Bhan AK. 1997. Suppressive role of B cells in chronic colitis of T cell receptor α mice. J. Exp. Med. 10:1749–56 Hermiston ML, Gordon JI. 1995. Inflammatory bowel disease and adenomas in mice expressing a dominant negative Ncadherin. Science 270:1203–7 Panwala CM, Jones JC, Viney JL. 1998. A novel model of inflammatory bowel disease: mice deficient for the multiple drug resistance gene, mdr1a, spontaneously develop colitis. J. Immunol. 161:5733–44 Madsen KL, Malfair D, Gray D, Doyle JS, Jewell LD, Fedorak RN. 1999. Inter-

67.

68. 69. 70.

71.

72.

73.

74.

75.

76.

leukin-10 gene-deficient mice develop a primary intestinal permeability defect in response to enteric microflora. Inflamm. Bowel Dis. 5:262–70 Mestecky J, Husby S, Moldoveanu Z, Waldo FB, van den Wall Bake AW, Elson CO. 1996. Induction of tolerance in humans: effectiveness of oral and nasal immunization routes. Ann. NY Acad. Sci. 778:194–201 Komagata Y, Weiner HL. 2000. Oral tolerance. Rev. Immunogenet. 2:61–73 Garside P, Mowat AM. 2001. Oral tolerance. Semin. Immunol. 13:177–85 Iwasaki A, Kelsall BL. 1999. Freshly isolated Peyer’s patch, but not spleen, dendritic cells produce interleukin 10 and induce the differentiation of T helper type 2 cells. J. Exp. Med. 190:229–39 Iwasaki A, Kelsall BL. 2001. Unique functions of CD11b+, CD8 alpha+, and double-negative Peyer’s patch dendritic cells. J. Immunol. 166:4884–90 Wirtz S, Finotto S, Lohse AW, Blessing M, Lehr HA, Galle PR, Neurath MF. 1999. Chronic intestinal inflammation in STAT-4 transgenic mice: characterization of disease and adoptive transfer by TNFplus IFN-γ -producing CD4+ T cells that respond to bacterial antigens. J. Immunol. 162:1884–88 Powrie F, Correa-Oliveira R, Mauze S, Coffman RL. 1994. Regulatory interactions between CD45RBhigh and CD45RBlow CD4+ T cells are important for the balance between protective and pathogenic cell-mediated immunity. J. Exp. Med. 179:589–600 Lucas PJ, Kim SJ, Melby SJ, Gress RE. 2000. Disruption of T cell homeostasis in mice expressing a T cell-specific dominant negative transforming growth factor beta II receptor. J. Exp. Med. 191:1187–96 Gorelik L, Flavell RA. 2000. Abrogation of TGFbeta signaling in T cells leads to spontaneous T cell differentiation and autoimmune disease. Immunity 12:171–81 Sartor RB. 1999. Microbial factors in

4 Feb 2002

11:44

AR

AR152-17.tex

AR152-17.SGM

LaTeX2e(2001/05/10)

P1: GKD

INSIGHTS INTO IBD

Annu. Rev. Immunol. 2002.20:495-549. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

77.

78.

79.

80.

81.

82.

83.

the pathogenesis of Crohn’s disease, ulcerative colitis, and experimental intestinal inflammation. In Inflammatory Bowel Diseases, ed. JB Kirsner, pp. 153–78. Philadelphia/London: Saunders. 5th ed. Sartor RB, Veltkamp C. 2000. Interactions between enteric bacteria and the immune system which determine mucosal homeostasis vs. chronic intestinal inflammation: lessons from rodent models. In IBD at the End of Its First Century, ed. G Rogler, F Kullman, P Rutgeerts, RB Sartor, J Scholmerich, pp. 30–41. Dordrecht, The Netherlands: Kluwer Academic Brimnes J, Reimann J, Nissen M, Claesson M. 2001. Enteric bacterial antigens activate CD4+ T cells from SCID mice with inflammatory bowel disease. Eur. J. Immunol. 31:23–31 Veltkamp C, Tonkonogy SL, De Jong YP, Albright C, Grenther WB, Balish E, Terhorst C, Sartor RB. 2001. Continuous stimulation by normal luminal bacteria is essential for the development and perpetuation of colitis in Tg(epsilon26) mice. Gastroenterology 120:900–13 Schultz M, Tonkonogy SL, Sellon RK, Veltkamp C, Godfrey VL, Kwon J, Grenther WB, Balish E, Horak I, Sartor RB. IL-2-deficient mice raised under germfree conditions develop delayed mild focal intestinal inflammation. Am. J. Physiol. Gastrointest. Liver Physiol. 276:G1461– 72 Contractor NV, Bassirir H, Reya T, Park AY, Baumgart DC, Wasik MA, Emerson SG, Carding SR. 1998. Lymphoid hyperplasia, autoimmunity, and compromised intestinal intraepithelial lymphocyte development in colitis-free gnotobiotic IL2-deficient mice. J. Immunol. 75:253– 61 Bhan AK, Mizoguchi E, Smith RN, Mizoguchi A. 2000. Spontaneous chronic colitis in TCR alpha-mutant mice; an experimental model of human ulcerative colitis. Int. Rev. Immunol. 19:123–38 Axelsson LG, Midtvedt T, Bylund-

84.

85.

86.

87.

88.

89.

90.

91.

541

Fellenius AC. 1996. The role of intestinal bacteria, bacterial translocation and endotoxin in dextran sodium sulfate-induced colitis in the mouse. Microb. Ecol. Health Dis. 9:225–37 Tlaskalova H, Stepankova R, Hudcovic T, Kozakova H. 1999. The role of bacterial microflora in development of dextran sodium sulfate (DSS) induced colitis in immunocompetent and immunodeficient mice. Microb. Ecol. Health Dis. 11:115– 16 (abstract) Madsen KL, Doyle JS, Tavernini MM, et al. 2000. Antibiotic therapy attenuates colitis in interleukin-10 gene deficient mice. Gastroenterology 118:1094–105 Sadlack B, Merz H, Schorle H, Schimpl A, Feller AC, Horak I. 1993. Ulcerative colitis-like disease in mice with a disrupted interleukin-2 gene. Cell 75:203–5 Duchmann R, Neurath MF, Meyer zum Buschenfelde KH. 1997. Responses to self and non-self intestinal microflora in health and inflammatory bowel disease. Res. Immunol. 148:589–94 Duchmann R, Schmitt E, Knolle P, Meyer zum Buschenfelde KH, Neurath M. 1996. Tolerance towards resident intestinal flora in mice is abrogated in experimental colitis and restored by treatment with interleukin-10 or antibodies to interleukin-12. Eur. J. Immunol. 26:934– 38 Ehrhardt RO, Ludviksson BR, Gray B, Neurath M, Strober W. 1997. Induction and prevention of colonic inflammation in IL-2-deficient mice. J. Immunol. 158:566–73 Duchmann R, May E, Heike M, Knolle P, Neurath M, Meyer zum Buschenfelde, KH. T cell specificity and cross reactivity towards enterobacteria, Bacterioides, bifidobacterium, and antigens from resident intestinal flora in humans. Gut 44:812–18 Cong Y, Weaver CT, Lazenby A, Elson CO. 2000. Colitis induced by enteric bacterial antigen-specific CD4+ T cells require CD40-CD40 ligand interactions for

4 Feb 2002

11:44

542

Annu. Rev. Immunol. 2002.20:495-549. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

92.

93.

94.

95.

96.

97.

98.

AR

AR152-17.tex

STROBER

¥

FUSS

AR152-17.SGM

¥

LaTeX2e(2001/05/10)

P1: GKD

BLUMBERG

a sustained increase in mucosal IL-12. J. Immunol. 165:2173–82 Cong Y, Brandwein SL, McCabe RP, Lazenby A, Birkenmeier EH, Sundberg JP, Elson CO. 1998. CD4+ T cells reactive to enteric bacterial antigens in spontaneously colitic C3H/HeJBir mice: increased T helper cell type 1 response and ability to transfer disease. J. Exp. Med. 16:855–64 Elson C, Cong Y, Iqbal N, Weaver C. 2001. Immuno-bacterial homeostasis in the gut: new insights into an old enigma. Semin. Immunol. 13:187–94 Aranda R, Sydora BC, McAllister PL, Binder SW, Yang HY, Targan SR, Kronenberg M. 1997. Analysis of intestinal lymphocytes in mouse colitis mediated by transfer of CD4+, CD45RBhigh T cells to SCID recipients. J. Immunol. 158:3464– 73 Probert CS, Chott A, Turner JR, Saubermann LJ, Stevens AC Bodinaku K, Elson CO, Balk SP, Blumberg RS. 1996. Persistent clonal expansion of peripheral blood CD4+ lymphocytes in chronic inflammatory bowel disease. J. Immunol. 157:3183–91 Sellon RK, Tonkonogy S, Schultz M, Duieleman LA, Grenther W, Balish E, Rennick DM, Sartor RB. 1998. Resident enteric bacteria are necessary for development of spontaneous colitis and immune system activation in interleukin-10deficient mice. Infect. Immun. 66:5244– 31 Kishi D, Takahashi I, Kai Y, Tamagawa H, Iijima H, Obunai S, Nezu R, Ito T, Matsuda H, Kiyono H. 2000. Alteration of V beta usage and cytokine production of CD4+ TCR beta beta homodimer T cells by elimination of Bacterioides vulgatus prevents colitis in TCR alpha-chaindeficient mice. J. Immunol. 165:5891– 99 Rodloff AC, Widera P, Ehlers S, et al. 1990. Suppression of blastogenic transformation of lymphocytes by Bacteri-

99.

100.

101.

102.

103.

104.

105.

106.

oides fragilis in vitro and in vivo. Int. J. Med. Microbiol. 274:406–16 Kullberg MC, Ward JM, Gorelick PL, Caspar P, Hieny S, Cheever A, Jankovic D, Sher A. 1998. Helicobacter hepaticus triggers colitis in specific-pathogenfree interleukin-10 (IL-10)-deficient mice through an IL-12 and gamma interferondependent mechanism. Infect. Immun. 66:5157–66 Dieleman LA, Arends A, Tonkonogy SL, Goerres MS, Craft DW, Grenther W, Sellon RK, Balish E, Sartor RB. Helicobacter hepaticus does not induce or potentiate colitis in interleukin-10-deficient mice. Infect. Immunol. 68:5107–13 von Freeden-Jeffrey U, Davidson N, Wiler R, Fort M, Burdach S, Murray R. 1998. IL-7 deficiency prevents development of a non-T cell non-B cell-mediated colitis. J. Immunol. 161:5673–80 Watanabe M, Ueno Y, Yajima T, Okamoto S, Hayashi T, Yamazaki M, Iwao Y, Ishii H, Haba S, Uehira M, Nishimoto H, Ishikawa H, Hata J, Hibi T. 1998. Interleukin 7 transgenic mice develop chronic colitis with decreased interleukin 7 protein accumulation in the colonic mucosa. J. Exp. Med. 187:389–402 Rath HC, Wilson KH, Sartor RB. 1999. Differential induction of colitis and gastritis in HLA-B27 transgenic rats selectively colonized with Bacterioides vulgatus or Escherichia coli. Infect. Immun. 67:2969– 74 Madsen KL, Doyle JS, Jewell LD, Tavernini MM, Fedorak RN. 1999. Lactobacillus species prevents colitis in interleukin 10 gene-deficient mice. Gastroenterology 116:1107–14 Maggie-Price L, Shows D, Waggie K, Burich A, Zeng W, Morrissey P, Viney JL. 2001. Immunologic responses in Helicobacter-induced inflammatory bowel disease in multiple drug resistance (mdr1a−/−) mice. Gastroenterology 120: 2667 (abstract) Dalwadi H, Wei B, Kronenberg M,

4 Feb 2002

11:44

AR

AR152-17.tex

AR152-17.SGM

LaTeX2e(2001/05/10)

P1: GKD

INSIGHTS INTO IBD

107.

Annu. Rev. Immunol. 2002.20:495-549. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

108.

109.

110.

111.

112.

113.

114.

115.

Sutton C, Braun J. 2001. The Crohn’s disease-associated bacterial protein I2 is a novel enteric T cell superantigen. Immunity 15:149–58 Mizoguchi A, Mizoguchi E, Chiba C, Bhan AK. 1996. Role of appendix in the development of inflammatory bowel disease in TCR-alpha mutant mice. J. Exp. Med. 184:707–15 Mahler M, Bristol IJ, Leiter EH, Workman AE, Birkenmeier EH, Elson CO, Sundberg JP. 1998. Differential susceptibility of inbred mouse strains to dextran sulfate sodium-induced colitis. Am. J. Physiol. 274:6544–51 Stevceva L, Pavli P, Buffinton G, Wozniak A, Doe WF. 1999. Dextran sodium sulphate-induced colitis activity varies with mouse strain but develops in lipopolysaccharide-unresponsive mice. J. Gastroenterol. Hepatol. 14:54–60 Mahler M, Bristol IJ, Sundberg JP, Churchill GA, Birkenmeier EH, Elson CO, Leiter EH. 1999. Genetic analysis of susceptibility to dextran sulfate sodiuminduced colitis in mice. Genomics 55: 147–56 Maehler M, Most C, Schmidtke S, Hedrich HJ. 2001. Genetic analysis of susceptibility to colitis in IL-10 deficient mice. Gastroenterology 120:185 (Abstract) Kozawa K, Sugawara K, Moskaluk CA, Matsumoto S, Cominelli F. 2001. Genetic analysis of susceptibility of ileitis in a spontaneous model of Crohn’s disease. Gastroenterology 120:186 (Abstract) Ahmad T, Satsangi J, McGovern D, Bunce M, Jewell DP. 2001. Review article: the genetics of inflammatory bowel disease. Aliment. Pharmacol. Ther. 18: 1937–41 Ogura Y, Bonen DK, Inohara N, Nicolae DL, Chen FF, Ramos R, et al. 2001. A frameshift mutation in NOD2 associated with susceptibility to Crohn’s disease. Nature 411:603–6 Hugot JP, Chamaillard M, Zouali H, Lesage S, Cezard JP, Belaiche J, et al.

116.

117.

118.

119.

120.

121.

122.

123.

543

2001. Association of NOD2 leucinerich repeat variants with susceptibility to Crohn’s disease. Nature 411:599–603 Beutler B. 2001. Autoimmunity and apoptosis: the Crohn’s connection. Immunity 15:1–14 Morris GP, Beck PL, Herridge MS, Depew WT, Szewczuk MR, Wallace JL. 1989. Hapten-induced model of colonic inflammation and ulceration in the rat colon. Gastroenterology 96:795–803 Yamada Y, Marshall S, Specian RD, Grisham MB. 1992. A comparative analysis of two models of colitis in rats. Gastroenterology 102:1524–34 St¨uber E, Strober W, Neurath M. 1996. Blocking the CD40L-CD40 interaction in vivo specifically prevents the priming of T helper 1 cells through the inhibition of interleukin 12 secretion. J. Exp. Med. 183:693–98 Sadlack B, Lohler J, Schorle H, Klebb G, Haber H, Sickel E, Noelle RJ, Horak I. 1995. Generalized autoimmune disease in interleukin-2-deficient mice is triggered by an uncontrolled activation and proliferation of CD4+ T cells. Eur. J. Immunol. 25:3055–59 De Jong YP, Comiskey M, Kalled SL, Mizoguchi E, Flavell RA, Bhan AK, Terhorst C. 2000. Chronic murine colitis is dependent on the CD154/CD40 pathway and can be attenuated by anti-CD154 administration. Gastroenterology 119:715– 23 Neurath MF, Fuss I, Pasparakis M, Alexopoulou L, Haralambous S, Meyer zum Buschenfelde KH, Strober W, Kollias G. 1997. Predominant pathogenic role of tumor necrosis factor in experimental colitis in mice. Eur. J. Immunol. 27:1743– 50 Fuss IJ, Marth T, Neurath M, Pearlstein GR, Jain A, Strober W. 1999. Anti-interleukin 12 treatment regulates apoptosis of Th1 T cells in experimental colitis in mice. Gastroenterology 117: 1078–88

4 Feb 2002

11:44

Annu. Rev. Immunol. 2002.20:495-549. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

544

AR

AR152-17.tex

STROBER

¥

FUSS

AR152-17.SGM

¥

LaTeX2e(2001/05/10)

P1: GKD

BLUMBERG

124. Neurath MF, Fuss IJ, Kelsall BL, Presky DH, Waegell W, Strober W. 1996. Experimental granulomatous colitis in mice is abrogated by induction of TGF-βmediated oral tolerance. J. Exp. Med. 183:2605–16 125. Elson CO, Beagley KW, Sharmanov AT, Fujihashi K, Kiyono H, Tennyson GS, Cong Y, Black CA, Ridwan BW, McGhee JR. 1996. Hapten-induced model of murine inflammatory bowel disease: mucosa immune response and protection by tolerance. J. Immunol. 157:2174–85 125a. Fuss IJ, Boirivant M, Lacey B, Strober W. 2002. The inter-related roles of TGFbeta and IL-10 in the regulation of experimental colitis. J. Immunol. In press 126. Powrie F, Leach MW, Mauze S, Caddle LB, Coffman RL. 1993. Phenotypically distinct subsets of CD4+ T cells induce or protect chronic intestinal inflammation in C.B-17 scid mice. Int. Immunol. 5:1461–71 127. Morrissey PJ, Charrier K, Braddy S, Liggitt D, Watson JD. 1993. CD4+ T cells that express high levels of CD45RB induce wasting disease when transferred into congenic severe combined immunodeficient mice. Disease development is prevented by cotransfer of purified CD4+ T cells. J. Exp. Med. 178:237–44 128. Powrie F, Leach MW, Mauze S, Menon S, Caddle LB, Coffman RL. 1994. Inhibition of Th1 responses prevents inflammatory bowel disease in scid mice reconstituted with CD45RBhi CD4+ T cells. Immunity 1:553–62 129. Leach MW, Bean AG, Mauze S, Coffman RL, Powrie F. 1996. Inflammatory bowel disease in C.B-17 scid mice reconstituted with the CD45RBhigh subset of CD4+ T cells. Am. J. Pathol. 148:1503– 15 130. Claesson MH, Bregenholt S, Bonhagen K, Thoma S, Moller P, Grusby MJ, Leithauser F, Nissen MH, Reimann J 1999. Colitis-inducing potency of CD4+ T

131.

132.

133.

134.

135.

136.

137.

138.

cells in immunodeficient, adoptive hosts depends on their state of activation, IL12 responsiveness, and CD45RB surface phenotype. J. Immunol. 162:3702–10 Matsuda JL, Gapin L, Sydora BC, Byrne F, Binder S, Kronenberg M, Aranda M. 2000. Systemic activation and antigen-driven oligoclonal expansion of T cells in a mouse model of colitis. J. Immunol. 164:2797–806 Powrie F, Carlino J, Leach MW, Mauze S, Coffman RL. 1996. A critical role for transforming growth factorbeta but not interleukin 4 in the suppression of T helper type 1-mediated colitis by CD45RBlow CD4+ T cells. J. Exp. Med. 183:2669–74 Read S, Malmstrom V, Powrie F. 2000. Cytotoxic T lymphocyte-associated antigen 4 plays an essential role in the function of CD25+CD4+ regulatory cells that control intestinal inflammation. J. Exp. Med. 192:295–302 Sakaguchi S, Toda M, Asano M, Itoh M, Morse SS, Sakaguchi N. 1996. T cellmediated maintenance of natural selftolerance: its breakdown as a possible cause of various autoimmune diseases. J. Autoimmun. 9:211–20 Bensinger SJ, Bandeira A, Jordon MS, Caton AJ, Laufer TM. 2001. Major histocompatibility complex class IIpositive cortical epithelium mediates the selection of CD4+25+ immunoregulatory T cells. J. Exp. Med. 194:427–38 Hagenbaugh A, Sharma S, Dubinett SM, Wei SH, Aranda R, Cheroutre H, et al. 1997. Altered immune responses in interleukin-10 transgenic mice. J. Exp. Med. 185:2101–10 Asseman C, Mauze S, Leach MW, Coffman RL, Powrie F. 1999. An essential role for interleukin 10 in the function of regulatory T cells that inhibit intestinal inflammation. J. Exp. Med. 190:995– 1004 Higgins LM, McDonald SA, Whittle N, Crockett N, Shields JG, MacDonald

4 Feb 2002

11:44

AR

AR152-17.tex

AR152-17.SGM

LaTeX2e(2001/05/10)

P1: GKD

INSIGHTS INTO IBD

Annu. Rev. Immunol. 2002.20:495-549. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

139.

140.

141.

142.

143.

144.

145.

146.

TT. 1999. Regulation of T cell activation in vitro and in vivo by targeting the OX40-OX40 ligand interaction: amelioration of ongoing inflammatory bowel disease with an OX40-IgG fusion protein, but not with an OX40 ligand-IgG fusion protein. J. Immunol. 162:486–93 Szabo SJ, Kim ST, Costa GL, Zhang X, Fathman CG, Glimcher LH. 2000. A novel transcription factor, T-bet, directs Th1 lineage commitment. Cell 100:665– 69 Mullen AC, High FA, Hutchins AS, Lee HW, Villarino AV, Livingston DM, Kung AL, Cereb N, Yao TP, Yang SY, Reiner SL. 2001. Role of T-bet in commitment of Th1 cells before IL-12 dependent selection. Science 292:1907–10 Jacobson NG, Szabo J, Weber-Nordt RM, Zhong Z, Schreiber D, Darnell JE, Murphy KM. 1995. Interleukin 12 signaling in T helper type 1 (Th1) cells involves tyrosine phosphorylation of signal transducer and activator of transcription (STAT)3 and STAT4. J. Exp. Med. 181:1755–62 Kaplan MH, Sun YL, Hoey T, Grusby MJ. 1996. Impaired IL-12 responses and enhanced development of Th2 cells in STAT4-deficient mice. Nature :382–174 Rudolph U, Finegold MJ, Rich SS, Harriman GR, Srinivasan Y, Brabet P, Bradley A, Birnbaumer L. 1995. Gi2 alpha protein deficiency: a model of inflammatory bowel disease. J. Clin. Immunol. 15:1015–55 He J, Gurunathan S, Iwasaki A, AshShaheed B, Kelsall BL. 2000. Primary role for Gi protein signaling in the regulation of interleukin 12 production and the induction of T helper cell type 1 responses. J. Exp. Med. 191:1605–10 Kuhn R, Lohler J, Rennick D, Rajewsky K, Muller W. 1993. Interleukin-10deficient mice develop chronic enterocolitis. Cell 75:263–74 Kullberg MC, Rothfuchs AG, Jankovic D, Caspar P, Wynn TA, Gorelick PL,

147.

148.

149.

150.

151.

152.

153.

545

Cheever AW, Sher A. 2001. Helicobacter hepaticus-induced colitis in interleukin-10-deficient mice: cytokine requirements for the induction and maintenance of intestinal inflammation. Infect. Immunol. 69:4232–41 Fiorentino DF, Zlotnick A, Mosmann TR, Howard M, O’Garra A. 1991. IL-10 inhibits cytokine production by activated macrophages. J. Immunol. 147:3815–22 Trinchieri G. 1997. Cytokines acting on or secreted by macrophages during intracellular infection (IL-10, IL-12, IFN-γ ). Curr. Opin. Immunol. 9:17–23 Ding L, Linsley PS, Huang LY, Germain RN, Shevach EM. 1993. IL-10 inhibits macrophage costimulatory activity by selectively inhibiting the upregulation of B7 expression. J. Immunol. 151:1224–34 Groux H, Bigler M, deVries JE, Roncarolo M. 1996. Interleukin-10 induces a long-term antigen-specific anergic state in human CD4+ T cells. J. Exp. Med. 184:19–29 Spencer SD, Di Marco F, Hooley J, Pitts-Meek S, Bauer M, Ryan AM, Sordat B, Gibbs VC, Aguet M. 1998. The orphan receptor CRF2-4 is an essential subunit of the interleukin 10 receptor. J. Exp. Med. 187:571–78 Suzuki A, Hanada T, Mitsuyama K, Takafumi Y, Kamizono S, Hoshino T, Kubo M, Yamashita A, Okabe M, Takeda K, Akira S, Matsumoto S, Toyonaga A, Sata M, Yoshimura A. 2001. CIS3/SOCS3/SSI3 plays a negative regulatory role in STAT3 activation and intestinal inflammation. J. Exp. Med. 193:471–81 Shull MM, Ormsby I, Kier AB, Pawloski S, Diebold RJ, Yin M, Allen R, Sidman C, Proetzel G, Calvin D, Annunziata N, Doetschman T. 1992. Targeted disruption of the mouse transforming growth factor-β1 gene results in multifocal inflammatory disease. Nature 359:693–99

4 Feb 2002

11:44

Annu. Rev. Immunol. 2002.20:495-549. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

546

AR

AR152-17.tex

STROBER

¥

FUSS

AR152-17.SGM

¥

LaTeX2e(2001/05/10)

P1: GKD

BLUMBERG

154. Kulkarni AB, Huh C-G, Becker D, Geiser A, Lyght M, Flanders KC, Roberts AB, Sporn MB, Ward JM, Karlsson S. 1993. Transforming growth factor β1 null mutation in mice causes excessive inflammatory response and early death. Proc. Natl. Acad. Sci. USA 90:770–74 155. Hahm KB, Im YH, Parks TW, Park SH, Markowitz S, Jung HY, Green J, Kim SJ. 2001. Loss of transforming growth factor beta signaling in the intestine contributes to tissue injury in inflammatory bowel disease. Gut 49:164–65 156. Simpson SJ, Shah S, Comiskey M, de Jong YP, Wang B, Mizoguchi E, Bhan AK, Terhorst C. 1998. T cell-mediated pathology in two models of experimental colitis depends predominantly on the interleukin 12/signal transducer and activator of transcription (STAT)-4 pathway, but is not conditional on interferon gamma expression by T cells. J. Exp. Med. 187:1225–34 157. Ludviksson BR, Ehrhardt RO, Strober W. 1997. TGF-beta production regulates the development of the 2,4,6-trinitrophenol-conjugated keyhold limpet hemocyanin-induced colonic inflammation in IL-2 deficient mice. J. Immunol. 159:3622–28 158. Hollander GA, Simpson SJ, Mizoguchi E, Nicogiannopoulou A, She J, Gutierrez-Ramos J, Bhan AK, Burakoff SJ, Wang B, Terhorst C. 1999. Severe colitis in mice with aberrant thymic selection. Immunity 3:27–38 159. Ludviksson BR, Gray B, Strober W, Ehrhardt RO. 1997. Dysregulated intrathymic development in the IL-2deficient mouse leads to colitis-inducing thymocytes. J. Immunol. 158:104–11 160. Kneitz B, Herrmann T, Yonehara S, Schimpl A. 1995. Normal clonal expansion but impaired Fas-mediated cell death and anergy induction in interleukin-2-deficient mice. Eur. J. Immunol. 25:2572–77 161. Razi-Wolf Z, Hollander GA, Reiser H.

162.

163.

164.

165.

166.

167.

168.

1996. Activation of CD4+ lymphocytes from interleukin-2-deficient mice by costimulatory B7 molecules. Proc. Natl. Acad. Sci. USA 93:2903–8 Telega GW, Baumgart DC, Carding SR. 2000. Uptake and presentation of antigen to T cells by primary colonic epithelial cells in normal and diseased states. Gastroenterology 119:1549–59 Willerford DM, Chen J, Ferry JA, Davidson L, Ma A, Alt FW. 1995. Interleukin-2 receptor α chain regulates the size and content of the peripheral lymphoid compartment. Immunity 3: 521–30 Poussier P, Ning T, Chen J, Banergee D, Julius M. 2000. Intestinal inflammation observed in IL-2R/IL-2 mutant mice is associated with impaired intestinal T lymphopoiesis. Gastroenterology 118:880–91 Suzuki H, Kundig T, Furlonger C, Wakeham A, Timms E, Matsuyama T, Schmits R, Simard J, Ohashi PS, Griesser H, Taniguchi T, Paige C, Mak TW. 1995. Deregulated T cell activation and autoimmunity in mice lacking interleukin-2 receptor β. Science 268:1472– 76 Wang B, Levelt C, Salio M, Zheng D, Sancho J, Liu CP, She J, Huang M, Higgins K, Sunshine MJ, et al. 1995. Over-expression of CD3 epsilon transgenes blocks T lymphocyte development. Int. Immunol. 7:435–58 Wang B, Simpson SJ, Hollander GA, Terhorst C. 1997. Development and function of T lymphocytes and natural killer cells after bone marrow transplantation of severely immunodeficient mice. Immunol. Rev. 157:53–60 Simpson SJ, Hollander GA, Mizoguchi E, Allen D, Bhan AK, Wang B, Terhorst C. 1997. Expression of proinflammatory cytokines by TCRαβ+ and TCRγ δ+ T cells in an experimental model of colitis. Eur. J. Immunol. 27:17– 25

4 Feb 2002

11:44

AR

AR152-17.tex

AR152-17.SGM

LaTeX2e(2001/05/10)

P1: GKD

Annu. Rev. Immunol. 2002.20:495-549. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

INSIGHTS INTO IBD 169. Sundberg JP, Elson CO, Bedigian H, Birkenmeier EH. 1994. Spontaneous heritable colitis in a new substrain of C3H/HeJ mice. Gastroenterology 107: 1726–35 170. Brandwein SL, McCabe RP, Cong Y, Waites KB, Ridwan BU, Dean PA, Ohkusa T, Birkenmeier EH, Sundberg JP, Elson CO. 1997. Spontaneously colitic C3H/HeJ Bir mice demonstrate selective antibody reactivity to antigens of the enteric bacterial flora. J. Immunol. 159:44–52 171. Sichler KJ, Overman KM, Sayed BA, Coroh B, Matsumoto S, Cominelli F, Pizarro TT. 2001. Dysregulated production of IEC-derived cytokines during early vs. established disease in the SAMP/Yit model of spontaneous ileitis. Gastroenterology 120:988 (abstract) 172. Mizoguchi A, Mizoguchi E, DeJong Y, Takedatsu H, Preffer FI, Terhorst C, Bhan AK. 2001. Role of CD5 expression on TCR γ δ expression on TCR γ δ T cell-differentiation and development of chronic intestinal inflammation. Gastroenterology 120:2629 (abstract) 173. Mizoguchi A, Mizoguchi E, Chiba C, Spiekermann GM, Tonegawa S, NaglerAnderson C, Bhan AK. 1996. Cytokine imbalance and autoantibody production in T cell receptor-alpha mutant mice with inflammatory bowel disease. J. Exp. Med. 183:847–56 174. Mizoguchi A, Mizoguchi E, Bhan AK. 1999. The critical role of interleukin 4 but not interferon gamma in the pathogenesis of colitis in T-cell receptor alpha mutant mice. Gastroenterology 116:320–26 175. Takashi I, Kiyono H, Hamada S. 1997. CD4+ T cell population mediates development of inflammatory bowel disease in T cell receptor alpha chain-deficient mice. Gastroenterology 112:1876–86 176. Iijima H, Takashi I, Kishi D, Kim JK, Kawano S, Hori M, Kiyono H. 1999. Alteration of interleukin 4 pro-

177.

178.

179.

180.

181.

182.

183.

184.

547

duction results in the inhibition of T helper type 2 cell-dominated inflammatory bowel disease in T cell receptor alpha chain-deficient mice. J. Exp. Med. 190:607–15 Coffman RL, Reiner SL. 1999. Instruction, selection, or tampering with the odds? Science 284:1283–85 Mombaerts P, Clarke AR, Rudnicki MA, Iacomini J, Itiohara S, Lafaille JJ, Wang L, Ichikawa Y, Jaenisch R, Hooper ML. 1992. Mutations in T cell antigen receptor genes alpha and beta block thymocytes development at different stages. Nature 360:225–31 Andres PG, Beck PL, Mizoguchi E, Mizoguchi A, Bhan AK, Dawson T, Kuziel WA, Maeda N, MacDermott RP, Podolsky DK, Reinecker HC. 2000. Mice with a selective deletion of the CC chemokine receptors 5 or 2 are protected from dextran sodium sulfate-mediated colitis: lack of CC chemokine receptor 5 expression results in a NK1.1+ lymphocyteassociated Th2-type immune response in the intestine. J. Immunol. 164:6303–12 Ranscht B. 1994. Cadherins and catenins: interactions and functions in embryonic development. Curr. Opin. Cell Biol. 6:740–46 Gottesman MM, Pastan I. 1993. Biochemistry of multidrug resistance mediated by the multidrug transporter. Annu. Rev. Biochem. 62:385–91 Leveille-Webster CR, Arias IM. 1995. The biology of the P-glycoprotein. J. Membr. Biol. 143:89–94 Satsangi J, Parkes M, Louis E, Hashimoto L, Kato N, Welsh K, Terwilliger JD, Lathrop GM, Bell JI, Jewell DP. 1996. Two stage genome-wide search inflammatory bowel disease provides evidence for susceptibility loci on chromosomes 3, 7, and 12. Nat. Genet. 14:199– 202 Yacyshyn B, Maksymowych W, Bowen-Yacyshyn MB. 1999. Differences in P-glycoprotein-170 expression

4 Feb 2002

11:44

548

185.

Annu. Rev. Immunol. 2002.20:495-549. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

186.

187.

188.

189.

190.

191.

192.

AR

AR152-17.tex

STROBER

¥

FUSS

AR152-17.SGM

¥

LaTeX2e(2001/05/10)

P1: GKD

BLUMBERG

and activity between Crohn’s disease and ulcerative colitis. Hum. Immunol. 60:677–87 Neish AS, Gewirtz AT, Zeng H, Young AN, Hobert ME, Karmali V, Rao AS, Madara JL. 2000. Prokaryotic regulation of epithelial responses by inhibition of IkappaB-alpha ubiquitination. Science 289:1560–63 Gewirtz AT, Navas TA, Lyons S, Godowski JP, Madara JL. 2001. Cutting edge: bacterial flagellin activates basolaterally expressed tlr5 to induce epithelial proinflammatory gene expression. J. Immunol. 167:1882–85 Egger B, Bajaj-Elliott M, MacDonald TT, Inglin R, Eysselein VE, Buchler MW. 2000. Characterization of acute murine dextran sodium sulphate colitis: cytokine profile and dose dependency. Digestion 62:240–48 Dielemann LA, Palmen MJ, Akol H, Bloemena E, Pena AS, Meuwissen SG, Van Rees EP. 1998. Chronic experimental colitis induced by dextran sulphate sodium (DSS) is characterized by Th1 and Th2 cytokines. Clin. Exp. Immunol. 114:385–91 Mashimo H, Wu DC, Podolsky DK, Fishman MC. 1996. Impaired defense of intestinal mucosa in mice lacking intestinal trefoil factor. Science 274:204 Panina-Bordignon P, Mazzeo D, Lucia PD, D’Ambrosio D, Lang R, Fabbri L, Self C, Sinigaglia F. 1997. Beta2agonists prevent Th1 development by selective inhibition of interleukin 12. J. Clin. Invest. 100:1513–19 Malfait AM, Malik AS, MarinovaMutafchieva L, Butler DM, Maini RN, Feldmann M. 1999. The beta2adrenergic agonist salbutamol is a potent suppressor of established collageninduced arthritis: mechanisms of action. J. Immunol. 162:6278–83 Van Assche G, Rutgeerts P. 2000. AntiTNF agents in Crohn’s disease. Expert Opin. Invest. Drugs. 9:103–11

193. Kishimoto T, Akira S, Taga T. 1992. Interleukin-6 and its receptor: a paradigm for cytokines. Science 258:593–97 194. Atreya R, Mudter J, Finotto S, Mullberg J, Jostock T, Wirtz S. et al. 2000. Blockade of interleukin 6 trans signaling suppresses T cell resistance against apopotosis in chronic intestinal inflammation: evidence in Crohn’s disease and experimental colitis in vivo. Nat. Med. 6:583–88 195. Barbara G, Xing Z, Hogaboam CM, Gauldie J, Collins SM. 2000. Interleukin 10 gene transfer prevents experimental colitis in rats. Gut 46:344–49 196. Tomoyose M, Mitsuyama K, Ishida H, Toyonaga A, Tanikawa K. 1998. Role of interleukin-10 in a murine model of dextran sulfate sodium-induced colitis. Scand. J. Gastroenterol. 33:435–40 197. Steidler L, Hans W, Schotte L, Neirynck S, Obermeier F, Falk W, Fiers W, Remaut E. 2000. Treatment of murine colitis by Lactococcus lactis secreting interleukin-10. Science 289:1352–55 198. Nelson DR, Lauwers GY, Lau JY, Davis GL. 2000. Interleukin 10 treatment reduces fibrosis in patients with chronic hepatitis C: a pilot trial of interferon nonresponders. Gastroenterology 118:655– 60 199. Picarella D, Hurlbut P, Rottman J, Shi X, Butcher E, Ringler DJ. 1997. Monoclonal antibodies specific for beta 7 integrin and mucosal addressin cell adhesion molecule-1 (MadCAM-1) reduce inflammation in the colon of SCID mice reconstituted with CD45RBhigh CD4+ T cells. J. Immunol. 158:2099–106 200. Kato S, Hokari R, Matsuzaki K, Iwai A, Kawaguchi A, Nagao S, Miyahara T, Itoh K, Ishii H, Miura S. 2000. Amelioration of murine experimental colitis by inhibition of mucosal addressin cell adhesion molecule-1. J. Pharmacol. Exp. Ther. 295:183–89 201. Ludviksson BR, Strober W, Nishikomori R, Hasan SK, Ehrhardt RO. 1999.

4 Feb 2002

11:44

AR

AR152-17.tex

AR152-17.SGM

LaTeX2e(2001/05/10)

P1: GKD

INSIGHTS INTO IBD

Annu. Rev. Immunol. 2002.20:495-549. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

202.

203.

204.

205.

206.

Administration of mAb against α Eβ7 prevents and ameliorates immunizationinduced colitis in IL-2−/− mice. J. Immunol. 162:4975–82 Herfarth H, Brand K, Rath K, Rogler HC, Scholmerich G, Falk W. 2000. Nuclear factor-kappa B activity and intestinal inflammation in dextran sulphate sodium (DSS)-induced colitis in mice is suppressed by gliotoxin. Clin. Exp. Immunol. 120:59–65 Hoffman MA, Drury S, Fu C, Qu W, Taguchi A, Lu Y, Avila C, Kambham N, Bierhaus A, Nawroth P, Neurath MF, Slattery T, Beach D, McClary J, Nagashima M, Morser J, Stern D, Schmidt AM. 1999. RAGE mediates a novel proinflammatory axis: a central cell surface receptor for S100/calgranulin polypeptides. Cell 97:889–901 Koh WP, Chan E, Scott K, McCaughan G, France M, Fazekas de St. Groth B. 1999. TCR-mediated involvement of CD4+ transgenic T cells in spontaneous inflammatory bowel disease in lymphopenic mice. J. Immunol. 162:7208– 16 Hammer RE, Maika SD, Richardson JA, Tang JP, Taurog JD. 1990. Spontaneous inflammatory disease in transgenic rats expressing HLA-B27 and human β2m: an animal model of HLAB27-associated human disorders. Cell 63:1099–112 Bertrand V, Quere S, Guimbaud R, Sogni P, Chauvelot-Moachon L, Tulliez M, Lamarque D, Charreire J, Giroud JP, Coutierier D, Chaussade S, Breban

207.

208.

209.

210.

211.

212.

549

M. 1998. Effects of murine recombinant interleukin-10 on the inflammatory disease of rats transgenic for HLA-B27 and human beta 2-microglobulin. Eur. Cytokine Netw. 9:161–70 Lee EG, Boone DL, Chai S, Libby SL, Chien M, Lodolce JP, Ma A. 2000. Failure to regulate TNF-induced NF-κB and cell death responses in A20-deficient mice. Science 289:2350–54 Beg AA, Sha WC, Bronson RT, Ghosh S, Baltimore D. 1995. Embryonic lethality and liver degeneration in mice lacking the Re1A component of NF-kappaB. Nature 376:167–70 Erdman S, Fox JG, Dangler CA, Feldman D, Horwitz BH. 2001. Typhlocolitis in NF-kappaB-deficient mice. J. Immunol. 166:1443–47 Dohi T, Fujihashi K, Rennert PD, Iwatani K, Kiyono H, McGhee JR. 1999. Hapten-induced colitis is associated with colonic patch hypertrophy and T helper cell 2-type responses. J. Exp. Med. 189:1169–80 Dohi T, Fujihashi K, Kiyono H, Elson CO, McGhee JR. 2000. Mice deficient in Th1- and Th2-type cytokines develop distinct forms of hapten-induced colitis. Gastroenterology 119:724–33 Snapper SB, Rosen FS, Mizoguchi E, Cohen P, Khan W, Liu C-H, Hagemann TL, Kwan SP, Ferrini R, Davidson L, Bhan AK, Alt FW. 1998. WiskottAldrich Syndrome protein-deficient mice reveal a role for WASP in T but not B cell activation. Immunity 9:81– 91

P1: FRK

February 1, 2002

12:47

Annual Reviews

AR152-FM

Annual Review of Immunology Volume 20, 2002

Annu. Rev. Immunol. 2002.20:495-549. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

CONTENTS FRONTISPIECE, Charles A. Janeway, Jr. A TRIP THROUGH MY LIFE WITH AN IMMUNOLOGICAL THEME, Charles A. Janeway, Jr.

xiv 1

THE B7 FAMILY OF LIGANDS AND ITS RECEPTORS: NEW PATHWAYS FOR COSTIMULATION AND INHIBITION OF IMMUNE RESPONSES, Beatriz M. Carreno and Mary Collins

29

MAP KINASES IN THE IMMUNE RESPONSE, Chen Dong, Roger J. Davis, and Richard A. Flavell

PROSPECTS FOR VACCINE PROTECTION AGAINST HIV-1 INFECTION AND AIDS, Norman L. Letvin, Dan H. Barouch, and David C. Montefiori T CELL RESPONSE IN EXPERIMENTAL AUTOIMMUNE ENCEPHALOMYELITIS (EAE): ROLE OF SELF AND CROSS-REACTIVE ANTIGENS IN SHAPING, TUNING, AND REGULATING THE AUTOPATHOGENIC T CELL REPERTOIRE, Vijay K. Kuchroo, Ana C. Anderson, Hanspeter Waldner, Markus Munder, Estelle Bettelli, and Lindsay B. Nicholson

55 73

101

NEUROENDOCRINE REGULATION OF IMMUNITY, Jeanette I. Webster, Leonardo Tonelli, and Esther M. Sternberg

125

MOLECULAR MECHANISM OF CLASS SWITCH RECOMBINATION: LINKAGE WITH SOMATIC HYPERMUTATION, Tasuku Honjo, Kazuo Kinoshita, and Masamichi Muramatsu

165

INNATE IMMUNE RECOGNITION, Charles A. Janeway, Jr. and Ruslan Medzhitov

KIR: DIVERSE, RAPIDLY EVOLVING RECEPTORS OF INNATE AND ADAPTIVE IMMUNITY, Carlos Vilches and Peter Parham ORIGINS AND FUNCTIONS OF B-1 CELLS WITH NOTES ON THE ROLE OF CD5, Robert Berland and Henry H. Wortis E PROTEIN FUNCTION IN LYMPHOCYTE DEVELOPMENT, Melanie W. Quong, William J. Romanow, and Cornelis Murre

197 217 253 301

LYMPHOCYTE-MEDIATED CYTOTOXICITY, John H. Russell and Timothy J. Ley

SIGNAL TRANSDUCTION MEDIATED BY THE T CELL ANTIGEN x

323

P1: FRK

Annu. Rev. Immunol. 2002.20:495-549. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

February 1, 2002

12:47

Annual Reviews

AR152-FM

CONTENTS

xi

RECEPTOR: THE ROLE OF ADAPTER PROTEINS, Lawrence E. Samelson INTERACTION OF HEAT SHOCK PROTEINS WITH PEPTIDES AND ANTIGEN PRESENTING CELLS: CHAPERONING OF THE INNATE AND ADAPTIVE IMMUNE RESPONSES, Pramod Srivastava CHROMATIN STRUCTURE AND GENE REGULATION IN THE IMMUNE SYSTEM, Stephen T. Smale and Amanda G. Fisher PRODUCING NATURE’S GENE-CHIPS: THE GENERATION OF PEPTIDES FOR DISPLAY BY MHC CLASS I MOLECULES, Nilabh Shastri, Susan

371

Schwab, and Thomas Serwold

395 427

463

THE IMMUNOLOGY OF MUCOSAL MODELS OF INFLAMMATION, Warren Strober, Ivan J. Fuss, and Richard S. Blumberg T CELL MEMORY, Jonathan Sprent and Charles D. Surh

GENETIC DISSECTION OF IMMUNITY TO MYCOBACTERIA: THE HUMAN MODEL, Jean-Laurent Casanova and Laurent Abel ANTIGEN PRESENTATION AND T CELL STIMULATION BY DENDRITIC CELLS, Pierre Guermonprez, Jenny Valladeau, Laurence Zitvogel, Clotilde Th´ery, and Sebastian Amigorena

495 551 581

621

NEGATIVE REGULATION OF IMMUNORECEPTOR SIGNALING, Andr´e Veillette, Sylvain Latour, and Dominique Davidson

669

CPG MOTIFS IN BACTERIAL DNA AND THEIR IMMUNE EFFECTS, Arthur M. Krieg

PROTEIN KINASE Cθ

709 IN

T CELL ACTIVATION, Noah Isakov and Amnon

Altman

RANK-L AND RANK: T CELLS, BONE LOSS, AND MAMMALIAN EVOLUTION, Lars E. Theill, William J. Boyle, and Josef M. Penninger PHAGOCYTOSIS OF MICROBES: COMPLEXITY IN ACTION, David M. Underhill and Adrian Ozinsky

761 795 825

STRUCTURE AND FUNCTION OF NATURAL KILLER CELL RECEPTORS: MULTIPLE MOLECULAR SOLUTIONS TO SELF, NONSELF DISCRIMINATION, Kannan Natarajan, Nazzareno Dimasi, Jian Wang, Roy A. Mariuzza, and David H. Margulies

853

INDEXES Subject Index Cumulative Index of Contributing Authors, Volumes 1–20 Cumulative Index of Chapter Titles, Volumes 1–20

ERRATA An online log of corrections to Annual Review of Immunology chapters (if any, 1997 to the present) may be found at http://immunol.annualreviews.org/

887 915 925

8 Feb 2002

10:3

AR

AR152-18.tex

AR152-18.SGM

LaTeX2e(2001/05/10) P1: GKD 10.1146/annurev.immunol.20.100101.151926

Annu. Rev. Immunol. 2002. 20:551–79 DOI: 10.1146/annurev.immunol.20.100101.151926 c 2002 by Annual Reviews. All rights reserved Copyright °

T CELL MEMORY Jonathan Sprent and Charles D. Surh Annu. Rev. Immunol. 2002.20:551-579. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

Department of Immunology, IMM4, The Scripps Research Institute, 10550 N. Torrey Pines Road, La Jolla, California 92037; e-mail: [email protected], [email protected]

Key Words T cells, memory, activation, survival, homeostasis, IL-15, IFN-I ■ Abstract Typical immune responses lead to prominent clonal expansion of antigen-specific T and B cells followed by differentiation into effector cells. Most effector cells die at the end of the immune response but some of these cells survive and form long-lived memory cells. The factors controlling the formation and survival of memory T cells are reviewed.

INTRODUCTION Exposure to infectious agents usually culminates in a state of immunological memory where secondary responses are more intense than primary responses (1–6). Memory is carried by antigen-specific T and B cells and is often lifelong. Here, we review recent information on memory T cells. The origin and differentiation of memory T cells and the factors controlling the long-term survival of these cells are discussed.

T CELLS AND THE PRIMARY IMMUNE RESPONSE Before considering how memory T cells arise during the primary immune response, it is useful to discuss some of the salient features of naive T cells in unstimulated animals.

Naive T Cells and the Preimmune Phase Prior to contact with antigen, naive T cells congregate in the secondary lymphoid tissues (spleen, lymph nodes, and Peyer’s patches) and migrate continuously from one lymphoid organ to another via blood and lymph (7–11). T cell migration through spleen is distinctly different from migration through lymph nodes and Peyer’s patches. Initial entry of T cells to the spleen is nonspecific. Like other lymphohemopoietic cells, T cells are carried into the spleen by the splenic artery and deposited in the marginal zone at the border between the red and white pulp. T (and B) cells 0732-0582/02/0407-0551$14.00

551

4 Feb 2002

11:50

Annu. Rev. Immunol. 2002.20:551-579. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

552

AR

AR152-18.tex

SPRENT

¥

AR152-18.SGM

LaTeX2e(2001/05/10)

P1: GKD

SURH

then move selectively to the white pulp where they accumulate around central arterioles; these areas, termed periarteriolar lymphocyte sheaths (PALS), are the main T cell zones in the spleen. How T cells move into PALS from the marginal zone is unclear, although it is notable that activated T cells lacking the chemokine receptor, CCR7, reach the red pulp but are excluded from the white pulp (12). Hence, T cell entry to the white pulp may be guided by CCR7 recognition of specific chemokines expressed on stromal cells. After about 12 h, naive T cells move from the PALS into the red pulp and leave the spleen via venous blood. In contrast to the spleen, entry to lymph nodes and Peyer’s patches from the blood is highly specific and occurs when cells enter high endothelial venules (HEV). For lymph nodes, the luminal surface of HEV expresses several ligands, including an addressin, PNAd (13), and a chemokine, SLC (14). These ligands are recognized by two lymph node homing receptors on T cells, namely CD62L and CCR7, respectively. Lymph node homing receptors enable T cells to bind to and penetrate the walls of HEV and reach the T cell zone (paracortex). Thereafter, T cells migrate out of the paracortex and leave lymph nodes via efferent lymphatic vessels, followed by re-entry into the bloodstream via the thoracic duct. This process of blood-to-lymph recirculation through lymph nodes takes about 12–18 hours (15). Continuous migration of T cells through the secondary lymphoid tissues is highly important for allowing T cells to make rapid contact with antigens released from pathogens. As discussed later, antigens are presented to T cells in the form of peptidic fragments bound to major histocompatibility complex (MHC) molecules (16). For naive T cells, these immunogenic peptide/MHC complexes have to be presented by specialized antigen-presenting cells (APC), especially by dendritic cells (17). These cells are strategically positioned as a dense network in the T cell zones and are continuously scrutinized by recirculating T cells for expression of foreign peptides. In unstimulated animals, presentation of peptide/MHC complexes by dendritic cells is limited to self peptides. Being largely tolerant of self components, naive T cells ignore the self-peptide/MHC complexes on normal dendritic cells, with the result that T cells are allowed to percolate slowly through the T cell zones and then re-enter the bloodstream for further recirculation. During their normal pattern of blood-to-lymph recirculation, naive T cells are metabolically quiescent and have a prolonged lifespan (18). Despite their inert appearance, however, the longevity of naive T cells is not innate but requires continuous contact with at least two external ligands, namely self-peptide/MHC complexes on dendritic cells (19–22) and a cytokine, IL-7 (23, 24). Recognition of these ligands presumably delivers low-level signals, which keep T cells sufficiently metabolically active to avoid passive death.

APC Activation and the Role of Adjuvants In considering how naive T cells first make contact with foreign antigens, it is important to stress that the migratory properties of naive T cells do not allow

4 Feb 2002

11:50

AR

AR152-18.tex

AR152-18.SGM

LaTeX2e(2001/05/10)

P1: GKD

Annu. Rev. Immunol. 2002.20:551-579. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

T CELL MEMORY

553

these cells to enter the initial site of infection, e.g., the lung for a respiratory infection. Naive T cells are programmed to recognize antigen only in the T cell zones of the secondary lymphoid tissues, and for this reason, initiation of the immune response depends upon transport of antigen to these areas from the site of infection. Antigens reach the T cell zones by two main routes: via afferent lymphatics for lymph nodes and via the bloodstream for the spleen. A priori, antigens could reach the T cell zones in soluble form and be degraded into peptides in situ by dendritic cells. This possibility is unlikely, however, because the mature dendritic cells found in the T cell zones are poorly equipped to ingest and process native proteins into peptides (25). By contrast, immature precursors of dendritic cells are highly efficient at antigen processing. In light of this finding, the current view is that, for the most part, antigens are conveyed to the T cell zones after phagocytosis by immature dendritic cell precursors (26–28). These cells are scattered throughout the body, including the skin (Langerhans cells), and ingest antigens at the site of infection. The antigen-laden cells then migrate to the secondary lymphoid tissues, differentiate into partly mature dendritic cells, and localize in the T cell zones; en route, the cells process antigen into immunogenic peptides and ferry these peptides to the cell surface on MHC molecules. This scheme is well documented for lymph nodes but is less clear for spleen. Since antigen-laden dendritic cell precursors entering the bloodstream are likely to be rapidly trapped by the liver and lungs, entry of these cells into the spleen is probably inefficient. A more likely scenario is that antigens enter the spleen in soluble form from the blood and are ingested by immature dendritic cell precursors situated in the marginal zone; these precursors then move into the white pulp and differentiate into mature dendritic cells in the PALS (29). In considering the movement of antigen-bearing cells into the T cell zones, it should be stressed that MHC-bound foreign peptides displayed on fully mature resting dendritic cells are poorly immunogenic (25). T cells do respond to these ligands, but the response is abortive and is often followed by the induction of tolerance, possibly because low expression of certain costimulatory molecules, e.g., B7 and/or CD40L, on resting dendritic cells prevents T cells from upregulating Bcl-2/Bcl-XL (30–32). For optimal immune responses, dendritic cells first need to be activated. This process is controlled by adjuvants. Most infectious microorganisms contain built-in adjuvants such as lipopolysaccharide (LPS), lipoproteins, unmethylated CpG DNA, and double-stranded viral RNA (33, 34). These products are recognized by cells of the innate immune system, including dendritic cells, by a spectrum of highly conserved, germ line–encoded Toll-like receptors (TLRs) (33–35). Through association with intracellular signaling molecules such as Myd88 (36, 37), ligation of TLRs by adjuvants then leads to cell activation. For immature dendritic cell precursors, such activation has several important consequences (26, 28, 38). First, antigen-laden immature dendritic cells are induced to leave mucosal sites and migrate to the T cell zones of draining lymph nodes; mobilization of dendritic cell precursors may be guided by a shift in the expression of chemokine receptors, CCR1 and CCR5 receptors for

4 Feb 2002

11:50

Annu. Rev. Immunol. 2002.20:551-579. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

554

AR

AR152-18.tex

SPRENT

¥

AR152-18.SGM

LaTeX2e(2001/05/10)

P1: GKD

SURH

inflammatory chemokines being replaced by CCR7 and other receptors for lymphoid chemokines (38). Second, stimulation of dendritic cells causes upregulation of essential costimulatory molecules for T cells such as B7-1 and B7-2, the ligands for CD28 (26). Third, dendritic cell activation results in synthesis of proinflammatory chemokines and cytokines (26, 28, 38); chemokines, such as SLC (6Ckine) may guide movement of activated CCR7+ dendritic cell precursors into the T cell zone, whereas APC release of cytokines, notably TNFα and IL-6, may provide soluble second signals for T cell activation (39, 40), thereby complementing costimulation provided by CD28/B7 interaction. In addition to acting directly on dendritic cells, some pathogens, especially viruses, can activate dendritic cells through production of type I (α, β) interferons (IFN-I) (41, 42). APC activation can also be induced by inflammatory molecules (TNFα and heat shock proteins) released through contact with necrotic cells (43). It should be noted that adjuvants are not needed for T cell responses to certain particulate antigens, e.g., heterologous erythrocytes (44), presumably because phagocytosis of these large antigens is sufficient to cause APC activation. Although the origin of activated mature dendritic cells from immigrant antigenloaded dendritic cell precursors is well accepted, the precise origin of dendritic cells is controversial (26, 28, 45). Currently, dendritic cells appear to arise from both myeloid and lymphoid precursors in blood; however, it is unclear whether these precursors represent two distinct lineages or reflect different pathways of development from a common precursor. As discussed elsewhere, myeloid and lymphoid dendritic cell precursors appear to be positioned differently but are both capable of transporting antigens to the T cell zones (46). Myeloid and lymphoid dendritic cells also show distinct differences in their capacity to synthesize certain cytokines, notably IL-12 and IFN-I [the latter being produced by plasmacytoid precursors in humans (41, 47)], and in their expression of individual TLRs (26, 28). Hence, selective activation of myeloid vs. lymphoid dendritic cell precursors by different pathogens may explain the considerable heterogeneity of T effector function seen in one infection vs. another.

T Cell Trapping by APC As discussed above, in the absence of foreign antigen, T cells move freely through the T cell zones, contact with self-peptides on dendritic cells being nonimmunogenic. When antigen-loaded APC enter the T cell zone, specifically reactive T cells recognize the foreign peptide/MHC complexes on these cells and within minutes form a tight synapse at the point of T/APC interaction (48–50). As discussed later, synapse formation is a crucial prelude for initiating T cell triggering and proliferation. However, the immediate consequence of T/APC synapse formation is that T cells become immobilized (44, 51). Such trapping in the T cell zones prevents blood-to-lymph recirculation and leads to continuous recruitment of newly arriving naive T cells into the immune response. Trapping is antigen-specific, and nonreactive T cells maintain their normal pattern of recirculation.

4 Feb 2002

11:50

AR

AR152-18.tex

AR152-18.SGM

LaTeX2e(2001/05/10)

P1: GKD

Annu. Rev. Immunol. 2002.20:551-579. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

T CELL MEMORY

555

Trapping of T cells in the lymphoid organs by APC is highly efficient. Thus, following intravenous (iv) injection of particulate antigens, such as heterologous erythrocytes in high doses, selective localization of these antigens in the spleen causes virtually all specifically reactive T cells to disappear from lymph nodes and thoracic duct lymph, and the T cells become trapped in the spleen (51). Such extensive trapping of T cells by the spleen after i.v. injection of antigen reflects the fact that the spleen has a large blood supply, which allows the vast majority of recirculating T cells to migrate through the spleen at least once a day. The situation is different for typical immune response induced by pathogens infecting mucosal sites. Here, via movement of activated antigen-laden APC, antigen is deposited largely in lymph nodes draining the sites of infection rather than in spleen. Since lymph nodes are normally small structures with a limited blood supply, only a tiny fraction of the total pool of recirculating T cells passes through an individual lymph nodes over a period of 24 hours. One might conclude therefore that recruitment of antigen-specific T cells to lymph nodes draining the site of infection is inefficient and limited to the few T cells that randomly enter lymph nodes from the blood. This may not be the case in practice, however, because draining lymph nodes become conspicuously enlarged within 1–2 days of infection (8). This phenomenon, originally misnamed lymph nodes shutdown, is poorly understood but is probably a manifestation of a sharp increase in the blood supply to infected lymph nodes, thus considerably accentuating the rate of perfusion of lymph nodes by recirculating T cells. In light of current knowledge (see above), the enhanced blood flow through infected lymph nodes may be mediated by chemokines released by the influx of activated antigen-laden APC. Whatever the explanation, a nonspecific increase in traffic through infected lymph nodes may be an important device for amplifying recruitment of antigen-specific naive T cells, thus increasing the precursor frequency of T cells during the early stages of the immune response.

T Cell Proliferation and Differentiation Because pathogens often replicate at a prodigious rate, clonal expansion of naive T cells and differentiation of these cells into effector cells has to be as rapid and efficient as possible and has to continue until the pathogen is eliminated, which takes 7–10 days for a typical viral infection. As mentioned above, initial interaction of T cells with APC leads to the formation of synapses at the T/APC contact site (48–50). Synapse formation precedes T cell activation and is associated with rapid clustering of TCR molecules binding to peptide/MHC complexes on APC plus local accumulation of intracellular signaling molecules such as LCK, LAT, and PCKθ ; the latter associate with TCR/CD3 complexes in lipid rafts and, together with various other intracellular molecules, initiate the downstream signaling events that cause T cells to proliferate, synthesize cytokines, and differentiate into effector cells. TCR/CD3 triggering is aided by CD4 and CD8 coreceptors and also by a large number of costimulatory/adhesion molecules on T cells (52–54). These molecules,

4 Feb 2002

11:50

Annu. Rev. Immunol. 2002.20:551-579. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

556

AR

AR152-18.tex

SPRENT

¥

AR152-18.SGM

LaTeX2e(2001/05/10)

P1: GKD

SURH

which include CD28, LFA-1, CD40L, ICOS, OX40, CD2, CD27, and 41BB, bind to complementary molecules on APC and, at least for CD28, CD2, and LFA-1, are drawn into the synapse along with TCR/CD3. Some costimulatory/adhesion molecules may provide essential second signals for T cell activation, but others may act largely by enhancing TCR triggering, e.g., by stabilizing synapse formation and/or recruiting intracellular signaling molecules (55, 56). Some costimulatory/adhesion molecules, notably CD28, are important for inducing cytokine (IL-2) synthesis by T cells, whereas others, e.g., CD40L, maintain or induce activation of APC and also stimulate B cells during T/B collaboration (57, 58). As mentioned earlier, optimal responses by naive T cells require APC activation by adjuvants. Through the release of stimulatory cytokines and upregulation of costimulating ligands, activated APC play a dual role in driving extensive T cell proliferation and promoting efficient differentiation into effector cells. However, the types of effector functions displayed by T cells following T/APC interaction appear to vary according to the stage of the immune response. During the early phase of the response, rapid replication of the pathogen ensures continuous entry of large numbers of activated APC into the T cell zones. Being highly stimulatory, these APC drive the responding T cells to proliferate rapidly (as often as every 8 h for CD8+ cells), synthesize a wide range of cytokines, and differentiate into cytotoxic cells (for CD8+ cells) and T helper cells (for CD4+ cells). Extensive clonal expansion applies to both CD4+ and CD8+ cells, although, at least in lymphocytic choriomeningitis virus (LCMV) infection, proliferation of CD4+ cells is somewhat slower and less prominent than for CD8+ cells (59–61). At later stages of the response, destruction of the pathogen at the site of infection by effector cells reduces inflow of antigen-laden APC into the T cell zones. Under these conditions, T cell interaction with diminishing numbers of antigenbearing APC exhausts the capacity of APC to produce stimulatory cytokines (62); the exhausted APC continue to elicit T cell proliferation but cytokine production by T cells and formation of fully differentiated effector cells are reduced. As discussed later, this stage of the immune response may be important for memory cell generation. The immune response presumably ceases when the influx of antigen-laden APC into the T cell zones is abolished. However, terminating the immune response may reflect the interplay of complex homeostatic mechanisms. For example, the recent finding that T cells can undergo up to seven rounds of cell division after being deprived of contact with APC (63, 64) may enable T cell proliferation and differentiation to continue for several days after the pathogen is cleared. Conversely, overexuberant T cell responses during the waning stages of the immune response may be limited by the action of inhibitory receptors on T cells, such as CTLA-4 and PD-1, binding to complementary ligands (B7 and PD-1L) on APC (65, 66). The notion that these receptors are inhibitory only when antigen becomes limiting is appealing—but unproven. With regard to effector cells, there is considerable interest in the range of cytokines produced by these cells and how cytokine synthesis is induced. Based on

4 Feb 2002

11:50

AR

AR152-18.tex

AR152-18.SGM

LaTeX2e(2001/05/10)

P1: GKD

Annu. Rev. Immunol. 2002.20:551-579. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

T CELL MEMORY

557

their patterns of cytokine production, T effector cells contain subsets of Th1 and Th2 cells (67). Th1 cells typically produce IFN-γ and TNFβ and protect against intracellular pathogens, whereas Th2 cells selectively synthesize IL-4, IL-5, IL-10, and IL-13 and counter extracellular pathogens through production of antibody by B cells. Other effector cells are uncommitted and either show limited cytokine synthesis or produce a nonpolarized spectrum of cytokines. The generation of Th1 vs. Th2 cells has been the topic of several recent reviews (28, 46, 68). For IFN-γ –producing Th1 cells, formation of these cells requires IL-12 production by APC. Since recently activated dendritic cells produce high levels of IL-12, IFN-γ production by T effector cells may be imprinted early in the immune response and continue until the response begins to wane. In viral infections, release of large amounts of IFN-I by APC may suppress IL-12 production (69). In this situation, IFN-I can substitute for IL-12 and lead to IFN-γ (and IL-10) synthesis by T cells via a STAT2/STAT4-dependent pathway (70); this pathway operates in humans but not in mouse because of a STAT2 defect (71). The generation of Th2 cells is less clear, but evidence is mounting that Th2 development may be a default pathway reflecting lack of exposure to IL-12 combined with contact with IL-4 (46). Th2 cells may be preferentially generated at the end of the immune response through contact with exhausted APC (62). During earlier stages of the response, high IL-12 synthesis by activated APC would be expected to preclude Th2 cell development. However, in this situation Th2 cells may arise through inhibition of IL-12 synthesis by suppressive cytokines, such as IL-10 and TGFβ (72, 73), and also by prostaglandin E2 (74).

FATE OF EFFECTOR CELLS In typical infections, prominent clonal expansion of specific T cells followed by differentiation into effector cells generally causes rapid elimination of pathogens. After elimination of the pathogen, the enormous numbers of effector cells become redundant and most of these cells rapidly disappear. The disappearance of activated T cells at the end of the primary response appears to reflect two distinct mechanisms, namely death and homing to nonlymphoid tissues. Early information on this issue came from studies on the fate of T cells responding to H2 alloantigens in vivo (75–77). Here, transferring parental-strain T cells to irradiated H2-heterozygous mice led to the appearance of large numbers of activated donor-derived effector T cells in thoracic duct lymph by 3–4 days posttransfer. When these effector cells were transferred to syngeneic (donor) mice, the cells homed initially to spleen and lymph nodes but then rapidly disappeared by two different mechanisms. First, many of the activated T cells left the spleen and localized in nonlymphoid tissues, notably the lungs, liver, and gut (75). Second, in both lymphoid and nonlymphoid tissues, the vast majority of the T cells died within 5 days (75), and only a small proportion of the cells survived to become

4 Feb 2002

11:50

558

AR

AR152-18.tex

SPRENT

¥

AR152-18.SGM

LaTeX2e(2001/05/10)

P1: GKD

SURH

long-lived functional memory cells (76, 77). Death was conspicuous in the liver— now regarded as a graveyard for dying T cells (78)–but was also apparent in the spleen; based on 51Cr-labeling studies, cells dying in the spleen were rapidly engulfed by phagocytic cells (75). Recent work on homing and death of effector T cells is discussed below.

Annu. Rev. Immunol. 2002.20:551-579. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

Homing of Effector Cells Detailed information on the capacity of effector T cells to migrate to nonlymphoid tissues has come from the tetramer staining method for detecting antigen-specific T cells in cell suspensions (79) and also by histological staining of T cells in whole-mouse sections (80). In these studies, clonal expansion of CD8+ and CD4+ T cells responding to viruses and soluble peptides led to the prominent appearance of activated effector T cells in various nonlymphoid tissues, including the lung, liver, gut, kidney, and salivary glands; unlike naive T cells, activated effector T cells also migrated to the bone marrow and the thymic medulla. In both lymphoid and nonlymphoid tissues, most of the effector T cells disappeared within 2–3 weeks, which presumably indicated death; however, some of the cells survived to form a subset of activated memory cells. The properties of these effector memory cells are considered later. As discussed elsewhere (11, 81), the capacity of activated T cells to migrate to nonlymphoid tissues is a reflection of upregulation of several classes of homing receptors, including selectins/selectin ligands, chemokine receptors, and both β1 and β2 integrins, notably LFA-1. Increased expression of these receptors allows activated T cells to bind to and penetrate the walls of small blood vessels and thereby percolate into extravascular sites throughout the body. To some extent, such migration is tissue-specific and reflects the particular homing receptors expressed. Thus, activated T cell homing to inflamed skin depends on binding to E/P-selectins and also to chemokines, e.g., TARC on skin vascular epithelium (82) and CTACK on keratinocytes (83); these molecules are recognized by complementary receptors on T cells, i.e., by E/P-selectin ligands and CCR4 (for TARC), respectively. Likewise, homing to the gut is controlled by α4β7 integrins and CCR9 on T cells recognizing MAdCAM-1 and chemokines (TECK), respectively, on gut vascular endothelium (81, 84). Homing to the peritoneal cavity involves E/P-selectin ligands and CXCR3 on T cells binding to E/P-selectins and IP10 on blood vessels (85). Within the secondary lymphoid tissues, migration of CD4+ T helper cells into B-cell follicles and germinal centers is controlled by CXCR5/BCA-1 interaction (86). In general, homing to nonlymphoid tissues is skewed toward cytokine-polarized effector cells, especially Th1 cells, and is largely restricted to T cells that have downregulated the lymph nodes-homing receptors, CD62L and CCR7 (12, 85, 87). The lack of these receptors prevents the effector cells from entering lymph nodes, but migration through the spleen is maintained. In spleen, the cells are found in the red pulp but are largely excluded from the white pulp (12).

4 Feb 2002

11:50

AR

AR152-18.tex

AR152-18.SGM

LaTeX2e(2001/05/10)

P1: GKD

T CELL MEMORY

559

Annu. Rev. Immunol. 2002.20:551-579. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

Death of Effector Cells Following early studies with superantigens (88), it is now well accepted that termination of the primary response is known to be followed by wide-scale death of effector cells. Teleologically, removal of effector cells after the pathogen is cleared makes sense because allowing these cells to survive en masse would overburden the immune system with useless cells: lymphoid hypertrophy and traffic congestion would ensue and thus compromise the response of naive T cells to new pathogens. This situation arises naturally in mice lacking certain cell-death– inducing molecules (see below). Here, unrestricted responses to environmental antigens lead to marked overproduction and survival of activated T cells, massive splenomegaly/lymphadenopathy occurs, and the mice die, presumably from infection. The mechanisms responsible for eliminating effector cells are poorly understood, but are clearly complex. Many cell-surface molecules on T cells, such as Fas and receptors for IL-2, TNF, and other cytokines, are able to transduce death signals under defined conditions (89–91), but the precise role of these molecules in mediating elimination of effector cells in vivo has yet to be resolved. For CD8+ cells, expansion/deletion of T cells responding to viral antigens is near normal in Bcl-2 and Bcl-XL transgenic mice (30, 92) and in mice lacking CTLA4 (93), Fas, or TNFR, or both Fas and TNFR (94, 95), which implies that these molecules are not needed for deletion. By contrast, IFN-γ seems to play a major role in the deletion of CD8+ cells. Thus, IFN-γ −/− mice show strong clonal expansion of CD8+ cells in response to viruses but very poor subsequent elimination of effector cells (96). With regard to other mediators, perforin appears to limit the intensity of the primary response of CD8+ cells (96, 97), perhaps through killing of APC, and may also control exhaustion of CD8+ cells responding to high doses of virus (96, and see below). However, perforin does not affect the deletion phase of the normal primary response (96). As for CD8+ cells, Fas/FasL interaction does not seem to be crucial for the elimination of effector CD4+ cells in normal immune responses (98, 99). However, Fas does control the deletion of CD4+ cells involved in chronic immune responses to self-antigens and perhaps also to continuously encountered environmental antigens (99). Unrestrained responses to these antigens are presumed to explain the prominent T cell hyperplasia seen in Fas (and FasL)-deficient mice (89). However, this syndrome is also conspicuous in a strikingly wide variety of gene-knockout mice, e.g., mice lacking PD-1 (66), CTLA4 (100, 101), NFAT (102), IL-2 (103), CD25 (104, 105), CD122 (106), CD45 (107), and TGFβ (108). Here, a key issue is whether the T cell hyperplasia in these knockout mice reflects defective immunoregulation of both CD4+ and CD8+ cells or only of CD4+ cells. For Fas deficiency, the effects of selectively depleting either CD4+ or CD8+ cells suggest that both subsets contribute to T cell hyperplasia (109, 110). For CTLA4−/− (111) and PD-1−/− mice (112), by contrast, disease onset is prevented by removal of CD4+ cells, which suggests that T hyperplasia is primarily under the control of

4 Feb 2002

11:50

Annu. Rev. Immunol. 2002.20:551-579. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

560

AR

AR152-18.tex

SPRENT

¥

AR152-18.SGM

LaTeX2e(2001/05/10)

P1: GKD

SURH

CD4+ cells; in these mice, recruitment of proliferating CD8+ cells could be a secondary event, perhaps reflecting enhanced exposure to stimulatory cytokines. Similarly, CD4+ cells rather than CD8+ cells seem to control T cell hyperplasia in IL-2−/− mice (113). In light of the above findings, the mechanisms controlling the elimination of effector CD4+ and CD8+ cells may be distinctly different. For CD8+ cells, the inhibitory influence of certain cytokines, notably IFN-γ , may be sufficient to eliminate effector cells. For CD4+ cells, by contrast, it would seem that effector cell elimination, which occurs more slowly than for CD8+ cells (60, 61), reflects a tightly regulated instructional process involving multiple cell-death–inducing mechanisms acting in consort; even with inactivation of only one of these mechanisms, death is averted and effector cells survive in large numbers. The precise sequence of molecular events required to destroy CD4+ effector cells is unclear but appears to involve negative signaling by CTLA-4 and PD-1 receptors for costimulatory molecules (66, 114), activation of the Fas death pathway by dissociation of cFLIP from Fas (115–117), and onset of sensitivity to several cytokines such as IL-2 (91), IFN-γ (118), and TNF (98, 119). In addition to these active mechanisms for cell death, effector cells may undergo passive death through loss of contact with protective cytokines (120, 121). This form of death can be prevented by enforced upregulation of Bcl-2 (99). Likewise, the finding that death of effector CD4+ cells is enhanced in CD40L−/− mice (59) suggests that passive death may also be promoted through lack of CD40L contact with CD40 on APC. Based on the above data, one can obviously discard the simple notion that effector cells are intrinsically short-lived cells doomed to die by a default pathway. Instead, effector cells have a propensity to survive, and eliminating these cells involves an instructional process. As discussed below, this conclusion has important implications for the generation of memory T cells.

SELECTION OF MEMORY T CELLS In typical primary responses, the wide-scale elimination of effector T cells at the end of the response is incomplete and a small proportion of T cells survive to become long-lived memory cells. Here, a key issue is whether production of T memory cells is a stochastic process or reflects selective mechanisms. On this point, it is important to consider whether T memory cells undergo affinity maturation.

Affinity Maturation In contrast to B cells there is no firm evidence that differentiation of naive T cells into memory cells reflects somatic hypermutation. However, bearing in mind that the TCR repertoire of naive T cells is highly diverse (122), one can envisage that contact with antigen during the primary immune response leads to significant alterations in the TCR specificity of memory cells relative to naive cells. In support of this notion, the TCR repertoire of T memory cells (CD4+ cells) is narrower than

4 Feb 2002

11:50

AR

AR152-18.tex

AR152-18.SGM

LaTeX2e(2001/05/10)

P1: GKD

Annu. Rev. Immunol. 2002.20:551-579. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

T CELL MEMORY

561

for naive cells (122, 123). It is also likely that T memory cells have a higher average affinity for antigen than do naive T cells (124–127). Here, an important question is whether such affinity maturation occurs solely during the primary response or can also proceed after the response is completed. It is easy to envisage affinity maturation during the primary response because, especially during the later stages of the response, high-affinity T cells are likely to have a growth advantage over low-affinity T cells because of competition for antigen (127). This mechanism is the simplest explanation for immunodominance during viral infections, i.e., stronger CD8+ T cell responses to dominant (strong) antigens than to subdominant (weak) antigens (128). In considering whether affinity maturation continues after the primary response, a key question is whether the repertoires of T memory cells and typical effector cells are the same or different. For CD8+ cells responding to pathogens, the bulk of evidence suggests that the repertoire of T memory cells is imprinted solely during the primary response; differences in the TCR repertoire of primary effector cells and long-lived memory cells are minimal (129, 130). This issue is less clear for CD4+ cells, although there are reports that affinity maturation of CD4+ cells can proceed for several weeks (123, 131). Whether these latter findings signify a fundamental difference between CD4+ and CD8+ cells is unclear. One possibility is that prolonged affinity maturation of CD4+ cells simply reflects a protracted primary response because of persistence of antigen, especially in germinal centers (123). The above data refer to memory cells tested directly ex vivo or after brief restimulation in vitro. When memory cells are restimulated in vivo, i.e., in secondary (recall) responses, clonal expansion of these cells can be skewed to high-affinity cells (129). However, this is not an invariable finding (130) and, when observed, this phenomenon may simply reflect enhanced competition for antigen in secondary responses because of rapid clearance of antigen by CTL or specific antibody.

Selection of Th1 and Th2 Memory Cells As discussed earlier, differentiation of T cells into polarized vs nonpolarized cells during the clonal expansion phase of the primary response is highly complex and reflects many different factors, including the density of antigen, TCR affinity, local contact with particular cytokines, and the nature of the APC. In general, polarized cells develop as the result of prolonged exposure to antigen, whereas nonpolarized cells arise by default when antigen is limiting (50, 87, 132). For memory cell generation, measuring cytokine profiles of memory T cells derived from precursors stimulated under polarizing conditions in vitro has shown that Th1 and Th2 effector populations retain their polarity for prolonged periods in vivo (3). Likewise, the particular cytokine profiles of primary effector T cells and long-term memory T cells are quite similar (3, 132). Such findings favor the view that selection of Th1 and Th2 effector cells for death vs. survival (as memory cells) is largely a stochastic process. It is worth noting, however, that the cytokine polarity of T cells is usually assessed after brief restimulation of T cells in vitro. Here, it is interesting that clonal analysis of cytokine production by CD4+ cells

4 Feb 2002

11:50

Annu. Rev. Immunol. 2002.20:551-579. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

562

AR

AR152-18.tex

SPRENT

¥

AR152-18.SGM

LaTeX2e(2001/05/10)

P1: GKD

SURH

tested directly ex vivo revealed distinct differences in the polarity of cells tested at day 9 of the primary response versus day 3 of the secondary response (132); these differences were largely masked when the cells were restimulated in vitro. In light of this finding, in terms of cytokine polarity the differentiation of effector cells into memory cells may to some extent involve selective mechanisms, at least for CD4+ cells. Despite the subtle differences observed between primary and secondary T cells in the above study (132), most of the T cells showed little evidence of cytokine polarity. This finding, which has also been reported by others (50, 87), is in line with the recent discovery that cytokine production and polarity of memory CD4+ T cells is largely restricted to a subset of activated cells marked by downregulation of CCR7 (50, 133). In contrast to these CCR7− T cells, the major population of “central” (resting) CCR7+ memory cells does not display cytokine polarity. As discussed below, this finding has important implications for memory cell selection.

Relationship of T Effectors and T Memory Cells Until recently, it was often argued that memory T cells could represent a subset of cells that failed to differentiate into effector cells during the primary response. This possibility now seems unlikely in view of evidence that the precursors of memory cells undergo extensive proliferation during the primary response (134–136) and, at least transiently, express effector functions such as perforin (136) and Granzyme B (137) synthesis and cytokine (IL-2) production (138). One has to conclude therefore that memory cells are derived from typical effector cells. What then is the difference between an effector cell that dies and an effector that differentiates into a memory cell? The simplest possibility is that there are no essential differences between typical T effectors and T memory precursors, generation of memory being largely a stochastic process (130). As suggested elsewhere (139, 140), an alternative possibility is that memory T cells are not drawn randomly from typical effector cells but are derived from a subset of precursors that arrives in the later stages of the immune response, e.g., because of slow recruitment to the draining lymph nodes from elsewhere in the body. These straggler cells may proliferate extensively and pass through an effector stage but, because of less-protracted contact with antigen, induction of the death pathways in these cells is incomplete or reversible; unlike terminally differentiated effectors, the straggler cells thus avoid death and survive to become memory cells. Though still hypothetical, this idea could explain the complete elimination of effector cells that occurs when T cells confront a very high dose of virus or other antigens (141, 142); here, continuous exposure to antigen causes exhaustive differentiation (death) of virtually all of the responding cells. For CD8+ cells, such death is partly avoided in perforin−/− mice (97). This finding may reflect that onset of CD8+ cell death is perforin-mediated (97). Alternatively, prolonged responses of normal CD8+ cells could lead to wide-scale perforin-dependent destruction of APC, T cell death being a consequence of abrupt loss of contact with antigen on APC (96, 143).

4 Feb 2002

11:50

AR

AR152-18.tex

AR152-18.SGM

LaTeX2e(2001/05/10)

P1: GKD

Annu. Rev. Immunol. 2002.20:551-579. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

T CELL MEMORY

563

The notion that T memory cells arise preferentially from straggler cells arriving late in the primary response has been invoked to explain the generation of nonpolarized central memory cells in humans (50). As discussed earlier, the suggestion here is that, during the terminal stages of the immune response, most APC express only low concentrations of antigen and are exhausted in terms of their capacity to produce polarizing cytokines, notably IL-12 for Th1 cells. Under these conditions, T cells recruited late in the response may avoid cytokine polarization and retain CCR7 expression; some cells die but others differentiate directly into CCR7+ longlived nonpolarized central memory cells. For cells recruited early in the primary response, by contrast, contact with high concentrations of antigen presented by activated APC plus cytokines generates CCR7− polarized effector cells, only a small proportion of which survive and differentiate into CCR7− polarized effector memory cells. Currently, the evidence that central memory cells are drawn from straggler precursors is largely indirect. In mice, useful information on this topic could be obtained by studying the efficiency of memory cell generation after exposing naive T cells to antigen-pulsed APC for various periods in vitro followed by adoptive transfer in vivo (63, 64). With this system, transfer of CD8+ cells after only brief (1 day) exposure to APC in vitro led to strong proliferation in vivo followed by differentiation into memory cells. Whether the extent of memory cell generation in this system is affected by the duration of in vitro stimulation and/or by antigen concentration and the activation status of the APC is still unclear. However, it is interesting that a recent study showed a highly efficient generation of memory cells when naive CD4+ cells were stimulated with antigen for 3 days in vitro, rested, and then adoptively transferred (144). Under these conditions of brief stimulation, the effector CD4+ cells evaded death and survived en masse to become memory cells. Viewed as a whole, the existing data indicate that memory cells arise from effector precursors, although it is not entirely clear whether these precursors represent fully differentiated or partly differentiated cells or a mixture of the two. The point to emphasize, however, is that memory cell generation appears to be an essentially passive process: Effector cells are coerced to die en masse by tightly regulated instructional mechanisms, and the few cells that escape death survive as memory cells, survival being a default pathway. This scheme applies to the initial generation of memory cells. As discussed later, the long-term survival of memory cells appears to involve instructional mechanisms.

Self-Specific Memory Cells Until recently, it was tacitly assumed that memory T cells are generated solely through contact with foreign antigens. In this respect, the sizeable proportion of T cells with a memory-phenotype (see below) found in normal animals is thought to reflect lifetime exposure to various environmental antigens. However, the assumption that memory cells are all descended from precursors responding to foreign antigens has been called into question by the recent finding that naive T cells begin to proliferate and differentiate into memory-phenotype

4 Feb 2002

11:50

564

AR

AR152-18.tex

SPRENT

¥

AR152-18.SGM

LaTeX2e(2001/05/10)

P1: GKD

SURH

Annu. Rev. Immunol. 2002.20:551-579. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

T cells when total numbers of naive T cells are reduced below a certain threshold (4, 145). This homeostatic response of T cells is directed largely to self antigens, especially to the particular self peptides that led to initial positive selection of naive T cells in the thymus. In view of this finding, the pool of memory T cells could be drawn in part from cells responding to self components. However, it should be emphasized that, except in the neonatal period and extreme old age, the pool size of naive T cells remains high, thus precluding homeostatic proliferation and differentiation. For this reason, the proportion of self-ligand–selected pseudo-memory cells is likely to be very low in normal adult animals.

FEATURES OF MEMORY CELLS Although T memory cells seem to survive almost indefinitely at a population level, these cells display considerable heterogeneity in terms of their surface markers, tissue distribution, and activation status.

Surface Markers The expression of surface markers on T memory cells has been extensively reviewed elsewhere (1–6). In general, memory cells show distinct phenotypic differences from naive T cells. Nevertheless, a number of the markers commonly used to define memory cells, e.g., low levels of CD62L and CD45RA/B/C, seem to be partly reversible on late memory cells (134, 146–149). However, it is unclear whether this reversion is real, or reflects preferential survival of cells that failed to lose these markers during the primary responses (150), although there is strong evidence for CD45RC− → CD45RC+ reversion in rats (151). This uncertainty also applies to CCR7 expression in humans (133). Thus, it has yet to be proved whether CCR7 expression on central memory cells reflects retention of CCR7 from the time of their differentiation from naive precursors or a CCR7+ → CCR7− → CCR7+ transition; in favor of this transition, day 8 effector cells generated in LCMV infection showed total absence of CCR7, though memory cells derived from the effectors were not tested (12). Currently, high expression of CD44 seems to be the most reliable marker for memory cells in mice, both for CD4+ and CD8+ cells; in addition, memory CD8+ cells are characterized by high expression of Ly6C and CD122 (150, 152). At a population level, memory cells show marked heterogeneity in their pattern of surface markers, with some cells resembling effector cells and others displaying many (but not all) of the characteristics of naive cells. Here, it is important to consider the activation status of memory cells.

Resting Versus Activated Cells Based on their rate of turnover (proliferation) in vivo and the expression of activation markers, memory T cells comprise two broad subsets of cells. Some memory cells are in an overtly activated state and closely resemble effector cells. These effector memory cells have a rapid turnover, show direct CTL

4 Feb 2002

11:50

AR

AR152-18.tex

AR152-18.SGM

LaTeX2e(2001/05/10)

P1: GKD

Annu. Rev. Immunol. 2002.20:551-579. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

T CELL MEMORY

565

activity (for CD8+ cells), and tend to express activation markers such as CD69 and CD25 (148, 149, 153–155). Like typical effector cells, activated memory cells express high levels of β1 and β2 integrins and specialized chemokine receptors that allow the cells to enter nonlymphoid tissues (133, 156). Activated memory cells are also found in the blood and spleen but are largely excluded from lymph nodes because of the loss of their lymph nodes homing receptors, CD62L and CCR7 (133); however, these cells do enter lymph nodes in small numbers via afferent lymphatic vessels (153). Resting (central) memory cells have a relatively slow turnover (148), lack activation markers, and closely resemble naive T cells in terms of their distribution in the lymphoid tissues (133). Central memory cells typically express CD62L and CCR7 receptors, which allows the cells to enter lymph nodes via high endothelial venules. Despite their quiescent state, resting memory cells are clearly less inert than naive T cells. Thus, for CD8+ cells, resting memory cells divide only intermittently but express quite high levels of RNA, which suggests that many of the cells are in the G1 phase of cell cycle (157); resting memory CD8+ cells also constitutively express perforin (152, 157, 158). The implication therefore is that memory cells are not totally quiescent but are maintained in a state of low-level activation. This topic is discussed later. The relationship between central and effector memory cells is unclear. The simplest idea is that these subsets represent two distinct populations derived from different precursors during the primary immune response. However, a more likely possibility is that activated memory cells arise after the primary response through stimulation of central memory cells (50, 133, 155), e.g., by cytokines, cross-reactive environmental antigens, or trace amounts of specific antigen. This issue could perhaps be resolved by studying the features of memory cells parked in MHC−/− hosts (see below).

Functions Despite the extensive culling of T cells at the end of the primary response, the precursor frequency of antigen-specific memory cells is far higher than for naive T cells (1–6, 59). This increase in precursor frequency is the main reason why secondary immune responses are generally much more intense than primary responses. In addition, it has long been argued that, cell-for-cell, memory T cells are more responsive to antigen than naive T cells. Because of the very low frequency of antigen-specific T cells in populations of normal naive T cells, a direct comparison of naive and memory cells is possible only with transgenic mice. Using this approach, it has been shown that populations enriched in resting memory T cells may (134) or may not (158, 159) show an increase in sensitivity to limiting concentrations of antigen (peptides) relative to naive T cells. However, when stimulated by antigen, memory cells do show a shorter lag time for entering cell cycle, synthesizing cytokines, differentiating into CTL, and migrating to nonlymphoid tissues (87, 134, 142, 149, 152, 157–159); memory cells are also somewhat less dependent on costimulation than naive T cells. These data are

4 Feb 2002

11:50

Annu. Rev. Immunol. 2002.20:551-579. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

566

AR

AR152-18.tex

SPRENT

¥

AR152-18.SGM

LaTeX2e(2001/05/10)

P1: GKD

SURH

consistent with the finding that memory cells are less metabolically inert than naive T cells. The above data apply to central memory cells. By contrast, effector memory cells, being in an overtly activated state, display constitutive CTL function and cytokine synthesis in vivo (50, 152, 158, 159). Residing in (or migrating through) nonlymphoid tissues such as the lung, effector memory cells are positionally poised to provide immediate immunity following secondary contact with the pathogen concerned (50, 154, 155). Secondary responses by central memory cells, by contrast, presumably hinge on initial reactivation by antigen in the draining lymph nodes before migration to nonlymphoid tissues. Because of this lag time, some workers argue that central memory cells are much less functionally relevant than effector memory cells (2). This point remains contentious.

MAINTENANCE The finding that memory cells are more metabolically active than naive T cells suggests that memory cells may be continuously signaled by extrinsic factors. Such signaling could be vital for keeping the cells alive. Here, two types of ligands could be involved.

Antigen In the past, it was argued that memory cells could be maintained through contact with trace amounts of specific antigen left over from the primary response (2). This idea now seems unlikely because memory T cells can survive in the apparent complete absence of specific antigen, both for CD8+ and CD4+ cells (21, 142, 157, 160–163). An alternative possibility is that memory cells receive survival signals through contact with cross-reactive environmental antigens (164). The obvious problem with this scenario is that it predicts that memory cells lacking crossreactivity for environmental antigens would rapidly disappear, for which there is little if any evidence. The only direct approach for testing whether memory cells are kept alive by contact with antigen is to follow the survival of these cells after transfer to MHC−/− hosts. Here, the clear-cut finding is that memory cells do survive for prolonged periods in MHC−/− hosts, both for CD4+ and CD8+ cells (144, 165, 166). However, it remains to be determined whether long-term survival in MHC−/− hosts applies to both effector and central memory cells or only to the latter. Nevertheless, at a population level, it would appear that the longevity of memory cells is independent of TCR ligation.

Cytokines As mentioned above, the turnover of memory cells, though slow relative to T cells proliferating during an immune response, is clearly faster than for naive T cells. Thus, whereas naive T cells rarely divide, memory cells divide intermittently, the period of interphase between divisions being in the order of 1–3 weeks (167). At

4 Feb 2002

11:50

AR

AR152-18.tex

AR152-18.SGM

LaTeX2e(2001/05/10)

P1: GKD

Annu. Rev. Immunol. 2002.20:551-579. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

T CELL MEMORY

567

least for CD8+ cells, this relatively rapid rate of cell division (compared to naive T cells) persists following transfer to MHC I−/− hosts (165). Evidence that cytokines affect the background turnover of CD8+ cells stemmed from experiments with agents that stimulate the innate immune system, e.g., Poly I:C and lipopolysaccharide (LPS) (168–171). Injecting these compounds into mice caused a brief burst of TCR-independent (bystander) proliferation, which was largely restricted to a subset of CD8+ cells with a memory (CD44hi) phenotype. Based on studies with IFN-γ −/− mice and mice lacking receptors for IFN-I, it was concluded that Poly I:C and LPS stimulate T cells via production of IFNs, both IFNI and IFN-γ . Since IFNs failed to stimulate purified T cells in vitro, it was reasoned that T cell proliferation in vivo reflects IFN-induced synthesis of another cytokine, an effector cytokine that acts directly on CD44hi CD8+ cells. IL-15 seemed a likely candidate because IFNs and IFN-inducing agents induced strong IL-15 mRNA synthesis by APC in vitro (172). In addition, CD122 (IL-2Rβ), a component of the receptor for IL-15 (and IL-2), was found to be expressed selectively on CD44hi CD8+ cells. The key finding was that IL-15 caused proliferation of purified CD44hi CD8+ cells (but not CD44hi CD4+ cells) in vitro and mimicked the capacity of IFNs to stimulate these cells in vivo. Recently, bystander proliferation of CD44hi CD8+ cells was found to be low or undetectable after transfer to IL-15−/− mice, which strongly implicates IL-15 as the effector cytokine (173). Bystander proliferation of CD44hi CD8+ cells is of brief duration and probably reflects augmentation of the natural turnover of these cells. This notion rests on the assumption that the relatively high turnover of CD44hi CD8+ cells is mediated through contact with background levels of IL-15. Strong support for this idea is provided by the finding that the normal turnover of CD44hi CD8+ cells in vivo is considerably reduced following injection of anti-CD122 mAb (174), presumably because this antibody blocks T cell contact with IL-15. It is also notable that total numbers of CD44hi CD8+ cells are selectively reduced in IL-15−/− mice (175) and increased in IL-15 transgenic mice (176, 177). Likewise, injection of IL-15 can boost numbers of antigen-specific memory CD8+ cells (178). In addition to providing a stimulus for cell division, IL-15 could also be important for keeping CD44hi CD8+ cells alive. The observation that CD44hi CD8+ cells are rare in IL-15−/− mice (175), and also in IL-15Rα −/− mice (179), is in favor of this possibility. In addition, CD44hi CD8+ cells rapidly disappear following transfer to IL-15−/− hosts (173). Collectively, the above data indicate that CD44hi CD8+ cells are strongly dependent on IL-15, both for their survival and turnover. It should be noted that dependency on IL-15 applies only to typical CD44hi CD8+ cells expressing a high density of CD122. A small proportion (about 30%) of CD44hi CD8+ cells are CD122lo, and these cells account for nearly all of the residual CD44hi CD8+ cells found in IL-15−/− mice (173). The implication therefore is that, unlike CD122hi cells, CD122lo CD44hi CD8+ cells are IL-15 independent. One recent study suggests that these latter cells could be short-lived. Thus, experiments in which CD8+ T cells were primed to specific antigen in vitro and then adoptively transferred in vivo showed that the fate of the transferred cells correlated with the level of CD122

4 Feb 2002

11:50

Annu. Rev. Immunol. 2002.20:551-579. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

568

AR

AR152-18.tex

SPRENT

¥

AR152-18.SGM

LaTeX2e(2001/05/10)

P1: GKD

SURH

expression: CD122hi cells survived long term, whereas CD122lo cells disappeared within a few weeks (180). Despite the conspicuous influence of IL-15 on CD8+ memory cells, CD4+ memory cells seem to be largely independent of IL-15. This finding is not surprising because CD122 expression on memory CD4+ cells is quite low (172). By analogy with CD8+ cells, it would seem quite likely that the survival of memory CD4+ cells is cytokine-dependent. However, it is striking that long-lived CD4+ memory cells can be generated from γ c−/− precursors (181), which implies that γ c-controlled cytokines (IL-2, -4, -7, -9, -15) are not important. Other cytokines have yet to be studied.

Homeostasis of Memory Cells Despite the fact that T memory cells have a high rate of turnover, total numbers of T memory cells remain relatively constant throughout life, though there are reports that numbers of CD4+ (60) and CD8+ (178) memory cells may eventually decline in old age. The implication therefore is that memory cells are subject to strict homeostatic control, background expansion of memory cells through intermittent cell division being countered by an equivalent level of cell death (182, 183). At a population level, this equilibrium presumably reflects a balance between lifesustaining signals and proapoptotic signals. As discussed above, the longevity of memory cells appears to depend on protective signals delivered by cytokines, notably by IL-15 for CD8+ cells. How cytokines promote memory cell survival is unclear, but upregulation of antiapoptotic molecules such as Bcl-2 and Bcl-XL is a likely possibility. Here, it is of interest that Bcl-2 upregulation in T cells in vitro is controlled by γ c-controlled cytokines (184), especially by IL-15 in CD8+ cells (185), whereas Bcl-XL upregulation is elicited by other cytokines, including IFN-I (186). Here, it is notable that when tested directly ex vivo, CD8+ memory cells show high levels of Bcl-2 and, to a lesser extent, Bcl-XL relative to naive T cells (185, 187). By contrast, CD4+ memory cells show upregulation of Bcl-XL but not Bcl-2 (142, 185). These findings add to the evidence that the mechanisms governing the survival of memory CD4+ and CD8+ cells are distinctly different (see above). Thus, memory CD8+ cells may be kept alive in vivo via Bcl-2 upregulation mediated by contact with IL-15 and other γ c cytokines, whereas memory CD4+ cells may be protected through Bcl-XL upregulation, e.g., by IFN-I. However, direct evidence on this issue is still limited, and upregulation of other pro-life molecules, e.g., LKLF (188, 189), could be equally important for maintaining memory T cell survival. Under steady-state conditions, memory T cell contact with life-sustaining cytokines is presumably limiting, thus curtailing continuous expansion of these cells. If so, a protracted increase in the level of these cytokines would be expected to cause numbers of memory T cells to increase above normal. The selective expansion of CD44hi CD8+ cells in IL-15 transgenic mice (176, 177) is consistent with this idea. These cells can also expand in response to IL-7. Thus, the marked

4 Feb 2002

11:50

AR

AR152-18.tex

AR152-18.SGM

LaTeX2e(2001/05/10)

P1: GKD

Annu. Rev. Immunol. 2002.20:551-579. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

T CELL MEMORY

569

increase in total T cell numbers seen in IL-7 transgenic mice (190) is strongly skewed to CD44hi CD8+ cells (W.C. Kieper, R. Ceredig, C.D. Surh, unpublished data). It is unclear whether other cytokine-transgenic mice, e.g., mice expressing high levels of IL-2 or IFN-I, display a similar phenotype. However, this approach is complicated by the fact that certain cytokines can stimulate the production of other cytokines, some of which could have either positive or negative effects on T cells. Here, it is notable that two γ c cytokines, IL-15 and IL-2, have opposing effects on CD44hi CD8+ cells. Thus, the background turnover of these cells is enhanced by IL-15 (see above) but inhibited by IL-2 (135, 174). Why IL-2 is inhibitory for memory CD8+ cells is unclear, but stimulation of a suppressive population of IL2–dependent regulatory CD4+ T cells is a possibility (191, 192). Nevertheless, a balance in the relative concentrations of IL-15 and IL-2 could be a key mechanism for controlling memory CD8+ cell homeostasis. In addition to IL-2, a number of other cytokines could have an inhibitory effect on memory cells. In this respect, it is striking that bystander activation of CD44hi CD8+ cells during viral infections can promote death of these cells and thus lead to attrition of memory (193); death is IFN-γ – and Fas-dependent and appears to reflect FasL+ antigen-specific T cells interacting with IFN-γ –conditioned bystander CD44hi CD8+ cells. Hence, in this situation, a direct inhibitory influence of IFN-γ on bystander cells counters the capacity of IFNs to protect these cells through production of IL-15. Such inhibition may contribute to the normal homeostasis of memory cells, but direct data on this issue are lacking. As discussed above, the survival of memory cells is regulated by complex homeostatic mechanisms. Precise information on how these mechanisms influence the life/death fate of memory cells, however, is still minimal. Clearly, much remains to be discovered in this important area. ACKNOWLEDGMENTS We thank Ms. Barbara Marchand for typing the manuscript. This work was supported by NIH grants CA38355, AI21487, AI46710, AG01743, AI41079, and AI45809. CDS is a Scholar of The Leukemia & Lymphoma Society. Publication no. 14307-IMM from the Scripps Research Institute. Visit the Annual Reviews home page at www.annualreviews.org

LITERATURE CITED 1. Ahmed R, Gray D. 1996. Immunological memory and protective immunity: understanding their relation. Science 272:54– 60 2. Zinkernagel RM, Bachmann MF, Kundig TM, Oehen S, Pirchet H, Hengartner H.

1996. On immunological memory. Annu. Rev. Immunol. 14:333–67 3. Dutton RW, Bradley LM, Swain SL. 1998. T cell memory. Annu. Rev. Immunol. 16:201–23 4. Goldrath AW, Bevan MJ. 1999. Selecting

4 Feb 2002

11:50

570

5.

6.

Annu. Rev. Immunol. 2002.20:551-579. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

7.

8.

9.

10.

11.

12.

13.

14.

15.

AR

AR152-18.tex

SPRENT

¥

AR152-18.SGM

LaTeX2e(2001/05/10)

P1: GKD

SURH

and maintaining a diverse T-cell repertoire. Nature 402:255–62 Lin MY, Selin LK, Welsh RM. 2000. Evolution of the CD8 T-cell repertoire during infections. Microbes. Infect. 2:1025–39 Sprent J, Surh CD. 2001. Generation and maintenance of memory T cells. Curr. Opin. Immunol. 13:248–54 Gowans JL, Knight EJ. 1964. The route of recirculation of lymphocytes in the rat. Proc. R. Soc. London Ser. B 159:257 Sprent J. 1977. Recirculating lymphocytes. In The Lymphocytes: Structure and Function, ed. JJ Marchalonis, pp. 43–112. New York: Dekker Picker LJ, Butcher EC. 1992. Physiological and molecular mechanisms of lymphocyte homing. Annu. Rev. Immunol. 10: 561–91 Sprent J. 1993. T lymphocytes and the thymus. In Fundamental Immunology, ed. WE Paul, pp. 75–110. New York: Raven. 3rd ed. Butcher EC, Picker LJ. 1996. Lymphocyte homing and homeostasis. Science 272:60–66 Potsch C, Vohringer D, Pircher H. 1999. Distinct migration patterns of naive and effector CD8 T cells in the spleen: correlation with CCR7 receptor expression and chemokine reactivity. Eur. J. Immunol. 29:3562–70 Berg EL, Robinson MK, Warnock RA, Butcher EC. 1991. The human peripheral lymph node vascular addressin is a ligand for LECAM-1, the peripheral lymph node homing receptor. J. Cell. Biol. 114:343– 49 Gunn MD, Tangemann K, Tam C, Cyster JG, Rosen SD, Williams LT. 1998. A chemokine expressed in lymphoid high endothelial venules promotes the adhesion and chemotaxis of naive T lymphocytes. Proc. Natl. Acad. Sci. USA 95:258– 63 Sprent J. 1973. Circulating T and B lymphocytes of the mouse. I. Migratory properties. Cell. Immunol. 7:10–39

16. Germain R. 1999. Antigen processing and presentation. In Fundamental Immunology, ed. WE Paul, pp. 287–340. Philadelphia: Raven. 4th ed. 17. Steinman RM. 1991. The dendritic cell system and its role in immunogenicity. Annu. Rev. Immunol. 9:271–96 18. Sprent J, Tough DF. 1994. Lymphocyte life-span and memory. Science 265:1395– 400 19. Takeda S, Rodewald H-R, Arakawa H, Bluethmann H, Shimizu T. 1996. MHC class II molecules are not required for survival of newly generated CD4+ T cells but affect their long-term life span. Immunity 5:217–28 20. Rooke R, Waltzinger C, Benoist C, Mathis D. 1997. Targeted complementation of MHC class II deficiency by intrathymic delivery of recombinant adenoviruses. Immunity 7:123–34 21. Tanchot C, Lemonnier FA, Perarnau B, Freitas AA, Rocha B. 1997. Differential requirements for survival and proliferation of CD8 naive or memory T cells. Science 276:2057–62 22. Ernst B, Lee D-S, Chang JM, Sprent J, Surh CD. 1999. The peptide ligands mediating positive selection in the thymus control T cell survival and homeostatic proliferation in the periphery. Immunity 11:173–81 23. Schluns KS, Kieper WC, Jameson SC, Lefrancois L. 2000. Interleukin-7 mediates the homeostasis of naive and memory CD8 T cells in vivo. Nature Immunol. 1:426–32 24. Tan JT, Dudl E, LeRoy E, Murray R, Sprent J, Weinberg KI, Surh CD. 2001. IL-7 is critical for homeostatic proliferation and survival of naive T cells. Proc. Natl. Acad. Sci. USA 98:8732–37 25. Jenkins MK, Khoruts A, Ingulli E, Mueller DL, McSorley SJ, Reinhardt RL, Itano A, Pape KA. 2001. In vivo activation of antigen-specific CD4 T cells. Annu. Rev. Immunol. 19:23–45 26. Banchereau J, Briere F, Caux C, Davoust

4 Feb 2002

11:50

AR

AR152-18.tex

AR152-18.SGM

LaTeX2e(2001/05/10)

P1: GKD

T CELL MEMORY

27.

Annu. Rev. Immunol. 2002.20:551-579. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

28.

29.

30.

31.

32.

33.

34.

35.

36.

J, Lebecque S, Liu YJ, Pulendran B, Palucka K. 2000. Immunobiology of dendritic cells. Annu. Rev. Immunol. 18:767– 811 Banchereau J, Steinman RM. 1998. Dendritic cells and the control of immunity. Nature 392:245–52 Liu YJ, Kanzler H, Soumelis V, Gilliet M. 2001. Dendritic cell lineage, plasticity and cross-regulation. Nat. Immunol. 2:585–89 De Smedt T, Pajak B, Muraille E, Lespagnard L, Heinen E, De Baetselier P, Urbain J, Leo O, Moser M. 1996. Regulation of dendritic cell numbers and maturation by lipopolysaccharide in vivo. J. Exp. Med. 184:1413–24 Petschner F, Zimmerman C, Strasser A, Grillot D, Nunez G, Pircher H. 1998. Constitutive expression of Bcl-xL or Bcl2 prevents peptide antigen-induced T cell deletion but does not influence T cell homeostasis after a viral infection. Eur. J. Immunol. 28:560–69 Bansal-Pakala P, Gebre-Hiwot Jember A, Croft M. 2001. Signaling through OX40 (CD134) breaks peripheral T-cell tolerance. Nat. Med. 7:907–12 Boise LH, Minn AJ, Noel PJ, June CH, Accavitti MA, Lindsten T, Thompson CB. 1995. CD28 costimulation can promote T cell survival by enhancing the expression of Bcl-xL. Immunity 3:87– 98 Aderem A, Ulevitch RJ. 2000. Toll-like receptors in the induction of the innate immune response. Nature 406:782–87 Akira S, Takeda K, Kaisho T. 2001. Toll-like receptors: critical proteins linking innate and acquired immunity. Nat. Immunol. 2:675–80 Medzhitov R, Preston-Hurlburt P, Janeway CA Jr. 1997. A human homologue of the Drosophila Toll protein signals activation of adaptive immunity. Nature 388:394–97 Muzio M, Ni J, Feng P, Dixit VM. 1997. IRAK (Pelle) family member IRAK-2 and

37.

38.

39.

40.

41.

42.

43.

44.

45.

46.

47.

571

MyD88 as proximal mediators of IL-1 signaling. Science 278:1612–15 Medzhitov R, Preston-Hurlburt P, Kopp E, Stadlen A, Chen C, Ghosh S, Janeway CA Jr. 1998. MyD88 is an adaptor protein in the hToll/IL-1 receptor family signaling pathways. Mol. Cell. 2:253–58 Sallusto F, Mackay CR, Lanzavecchia A. 2000. The role of chemokine receptors in primary, effector, and memory immune responses. Annu. Rev. Immunol. 18:593– 620 Kishimoto H, Sprent J. 1999. Strong TCR ligation without costimulation causes rapid onset of Fas-dependent apoptosis of naive murine CD4+ T cells. J. Immunol. 163:1817–26 Sepulveda H, Cerwenka A, Morgan T, Dutton RW. 1999. CD28, IL-2-independent costimulatory pathways for CD8 T lymphocyte activation. J. Immunol. 163:1133–42 Siegal FP, Kadowaki N, Shodell M, Fitzgerald-Bocarsly PA, Shah K, Ho S, Antonenko S, Liu YJ. 1999. The nature of the principal type 1 interferon-producing cells in human blood. Science 284:1835– 37 Biron CA. 2001. Interferons alpha and beta as immune regulators—a new look. Immunity 14:661–64 Gallucci S, Matzinger P. 2001. Danger signals: SOS to the immune system. Curr. Opin. Immunol. 13:114–19 Sprent J. 1978. Role of H-2 gene products in the function of T helper cells from normal and chimeric mice in vivo. Immunol. Rev. 42:108–37 Shortman K. 2000. Burnet oration: dendritic cells: multiple subtypes, multiple origins, multiple functions. Immunol. Cell. Biol. 78:161–65 Moser M, Murphy KM. 2000. Dendritic cell regulation of TH1-TH2 development. Nat. Immunol. 1:199–205 Cella M, Facchetti F, Lanzavecchia A, Colonna M. 2000. Plasmacytoid dendritic cells activated by influenza virus and

4 Feb 2002

11:50

572

48.

Annu. Rev. Immunol. 2002.20:551-579. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

49.

50.

51.

52.

53.

54.

55.

56.

57.

AR

AR152-18.tex

SPRENT

¥

AR152-18.SGM

LaTeX2e(2001/05/10)

P1: GKD

SURH

CD40L drive a potent TH1 polarization. Nat. Immunol. 1:305–10 Monks CR, Freiberg BA, Kupfer H, Sciaky N, Kupfer A. 1998. Threedimensional segregation of supramolecular activation clusters in T cells. Nature 395:82–86 Grakoui A, Bromley SK, Sumen C, Davis MM, Shaw AS, Allen PM, Dustin ML. 1999. The immunological synapse: a molecular machine controlling T cell activation. Science 285:221–27 Lanzavecchia A, Sallusto F. 2000. Dynamics of T lymphocyte responses: intermediates, effectors, and memory cells. Science 290:92–97 Sprent J, Miller JF. 1974. Effect of recent antigen priming on adoptive immune responses. II. Specific unresponsiveness of circulating lymphocytes from mice primed with heterologous erythrocytes. J. Exp. Med. 139:1–12 Chambers CA, Allison JP. 1999. Costimulatory regulation of T cell function. Curr. Opin. Cell. Biol. 11:203–10 Watts TH, DeBenedette MA. 1999. T cell co-stimulatory molecules other than CD28. Curr. Opin. Immunol. 11:286–93 Coyle AJ, Gutierrez-Ramos JC. 2001. The expanding B7 superfamily: increasing complexity in costimulatory signals regulating T cell function. Nat. Immunol. 2:203–9 Bromley SK, Burack WR, Johnson KG, Somersalo K, Sims TN, Sumen C, Davis MM, Shaw AS, Allen PM, Dustin ML. 2001. The immunological synapse. Annu. Rev. Immunol. 19:375–96 Lanzavecchia A, Sallusto F. 2001. Antigen decoding by T lymphocytes: from synapses to fate determination. Nat. Immunol. 2:487–92 June CH, Ledbetter JA, Gillespie MM, Lindsten T, Thompson CB. 1987. T-cell proliferation involving the CD28 pathway is associated with cyclosporine-resistant interleukin 2 gene expression. Mol. Cell. Biol. 7:4472–81

58. Walker LS, Gulbranson-Judge A, Flynn S, Brocker T, Lane PJ. 2000. Co-stimulation and selection for T-cell help for germinal centres: the role of CD28 and OX40. Immunol. Today 21:333–37 59. Whitmire JK, Ahmed R. 2000. Costimulation in antiviral immunity: differential requirements for CD4+ and CD8+ T cell responses. Curr. Opin. Immunol. 12:448– 55 60. Homann D, Teyton L, Oldstone MB. 2001. Differential regulation of antiviral T-cell immunity results in stable CD8+ but declining CD4+ T-cell memory. Nat. Med. 7:913–19 61. Whitmire JK, Ahmed R. 2001. The economy of T-cell memory: CD4+ recession in times of CD8+ stability? Nat. Med. 7:892– 93 62. Langenkamp A, Messi M, Lanzavecchia A, Sallusto F. 2000. Kinetics of dendritic cell activation: impact on priming of TH1, TH2 and nonpolarized T cells. Nat. Immunol. 1:311–16 63. Kaech SM, Ahmed R. 2001. Memory CD8+ T cell differentiation: initial antigen encounter triggers a developmental program in naive cells. Nat. Immunol. 2:415– 22 64. van Stipdonk MJ, Lemmens EE, Schoenberger SP. 2001. Naive CTLs require a single brief period of antigenic stimulation for clonal expansion and differentiation. Nat. Immunol. 2:423–29 65. Walunas TL, Lenschow DJ, Bakker CY, Linsley PS, Freeman GJ, Green JM, Thompson CB, Bluestone JA. 1994. CTLA-4 can function as a negative regulator of T cell activation. Immunity 1:405– 13 66. Freeman GJ, Long AJ, Iwai Y, Bourque K, Chernova T, Nishimura H, Fitz LJ, Malenkovich N, Okazaki T, Byrne MC, Horton HF, Fouser L, Carter L, Ling V, Bowman MR, Carreno BM, Collins M, Wood CR, Honjo T. 2000. Engagement of the PD-1 immunoinhibitory receptor by a novel B7 family member leads

4 Feb 2002

11:50

AR

AR152-18.tex

AR152-18.SGM

LaTeX2e(2001/05/10)

P1: GKD

T CELL MEMORY

67.

68.

Annu. Rev. Immunol. 2002.20:551-579. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

69.

70.

71.

72.

73.

74.

75.

76.

to negative regulation of lymphocyte activation. J. Exp. Med. 192:1027–34 Mosmann TR, Sad S. 1996. The expanding universe of T-cell subsets: Th1, Th2 and more. Immunol. Today 17:138–46 Reiner SL. 2001. Helper T cell differentiation, inside and out. Curr. Opin. Immunol. 13:351–55 Cousens LP, Orange JS, Su HC, Biron CA. 1997. Interferon-alpha/beta inhibition of interleukin 12 and interferongamma production in vitro and endogenously during viral infection. Proc. Natl. Acad. Sci. USA 94:634–39 Farrar JD, Smith JD, Murphy TL, Murphy KM. 2000. Recruitment of Stat4 to the human interferon-alpha/beta receptor requires activated Stat2. J. Biol. Chem. 275:2693–97 Farrar JD, Smith JD, Murphy TL, Leung S, Stark GR, Murphy KM. 2000. Selective loss of type I interferon-induced STAT4 activation caused by a minisatellite insertion in mouse Stat2. Nat. Immunol. 1:65–69 Moore KW, de Waal Malefyt R, Coffman RL, O’Garra A. 2001. Interleukin10 and the interleukin-10 receptor. Annu. Rev. Immunol. 19:683–765 King C, Davies J, Mueller R, Lee MS, Krahl T, Yeung B, O’Connor E, Sarvetnick N. 1998. TGF-beta1 alters APC preference, polarizing islet antigen responses toward a Th2 phenotype. Immunity 8:601– 13 Kalinski P, Vieira PL, Schuitemaker JH, de Jong EC, Kapsenberg ML. 2001. Prostaglandin E(2) is a selective inducer of interleukin-12 p40 (IL-12p40) production and an inhibitor of bioactive IL-12p70 heterodimer. Blood 97:3466–69 Sprent J. 1976. Fate of H-2-activated T lymphocytes in syngeneic hosts. I. Fate in lymphoid tissues and intestines traced with 3H-thymidine, 125I-deoxyuridine and 51 chromium. Cell. Immunol. 21:278– 302 Sprent J, Miller JF. 1976. Fate of H2-

77.

78.

79.

80.

81.

82.

83.

84.

573

activated T lymphocytes in syngeneic hosts. II. Residence in recirculating lymphocyte pool and capacity to migrate to allografts. Cell Immunol. 21:303–13 Sprent J, Miller JFAP. 1976. Fate of H2-activated T lymphocytes in syngeneic hosts. III. Differentiation into long-lived recirculating memory cells. Cell. Immunol. 21:314–26 Huang L, Soldevila G, Leeker M, Flavell R, Crispe IN. 1994. The liver eliminates T cells undergoing antigen-triggered apoptosis in vivo. Immunity 1:741–49 Masopust D, Vezys V, Marzo AL, Lefrancois L. 2001. Preferential localization of effector memory cells in nonlymphoid tissue. Science 291:2413–17 Reinhardt RL, Khoruts A, Merica R, Zell T, Jenkins MK. 2001. Visualizing the generation of memory CD4 T cells in the whole body. Nature 410:101–5 Campbell JJ, Butcher EC. 2000. Chemokines in tissue-specific and microenvironment-specific lymphocyte homing. Curr. Opin. Immunol. 12:336–41 Campbell JJ, Haraldsen G, Pan J, Rottman J, Qin S, Ponath P, Andrew DP, Warnke R, Ruffing N, Kassam N, Wu L, Butcher EC. 1999. The chemokine receptor CCR4 in vascular recognition by cutaneous but not intestinal memory T cells. Nature 400:776–80 Morales J, Homey B, Vicari AP, Hudak S, Oldham E, Hedrick J, Orozco R, Copeland NG, Jenkins NA, McEvoy LM, Zlotnik A. 1999. CTACK, a skinassociated chemokine that preferentially attracts skin-homing memory T cells. Proc. Natl. Acad. Sci. USA 96:14,470–75 Zabel BA, Agace WW, Campbell JJ, Heath HM, Parent D, Roberts AI, Ebert EC, Kassam N, Qin S, Zovko M, LaRosa GJ, Yang LL, Soler D, Butcher EC, Ponath PD, Parker CM, Andrew DP. 1999. Human G protein-coupled receptor GPR-9-6/CC chemokine receptor 9 is selectively expressed on intestinal homing T lymphocytes, mucosal lymphocytes, and

4 Feb 2002

11:50

574

85.

Annu. Rev. Immunol. 2002.20:551-579. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

86.

87.

88.

89. 90.

91.

92.

93.

94.

AR

AR152-18.tex

SPRENT

¥

AR152-18.SGM

LaTeX2e(2001/05/10)

P1: GKD

SURH

thymocytes and is required for thymusexpressed chemokine-mediated chemotaxis. J. Exp. Med. 190:1241–56 Xie H, Lim YC, Luscinskas FW, Lichtman AH. 1999. Acquisition of selectin binding and peripheral homing properties by CD4+ and CD8+ T cells. J. Exp. Med. 189:1765–76 Schaerli P, Willimann K, Lang AB, Lipp M, Loetscher P, Moser B. 2000. CXC chemokine receptor 5 expression defines follicular homing T cells with B cell helper function. J. Exp. Med. 192:1553– 62 Iezzi G, Scheidegger D, Lanzavecchia A. 2001. Migration and function of antigenprimed nonpolarized T lymphocytes in vivo. J. Exp. Med. 193:987–93 Sprent J, Webb SR. 1995. Intrathymic and extrathymic clonal deletion of T cells. Curr. Opin. Immunol. 7:196–205 Nagata S. 1997. Apoptosis by death factor. Cell 88:355–65 Rathmell JC, Thompson CB. 1999. The central effectors of cell death in the immune system. Annu. Rev. Immunol. 17:781–828 Leonardo M, Chan FKM, Hornung F, McFarland H, Siegel R, Wang J, Zheng L. 1999. Mature T lymphocyte apoptosisimmune regulation in a dynamic and unpredictable antigenic environment. Annu. Rev. Immunol. 17:221–53 Razvi ES, Jiang Z, Woda BA, Welsh M. 1995. Lymphocyte apoptosis during the silencing of the immune response to acute viral infections in normal, lpr, and Bcl-2 transgenic mice. Am. J. Pathol. 147:79– 91 Bachmann MF, Waterhouse P, Speiser DE, McKall-Faienza K, Mak TW, Ohashi PS. 1998. Normal responsiveness of CTLA-4-deficient anti-viral cytotoxic T cells. J. Immunol. 160:95–100 Reich A, Korner H, Sedgwick JD, Pircher H. 2000. Immune down-regulation and peripheral deletion of CD8 T cells does not require TNF receptor-ligand in-

95.

96.

97.

98.

99.

100.

101.

102.

103.

teractions nor CD95 (Fas, APO-1). Eur. J. Immunol. 30:678–82 Nguyen LT, McKall-Faienza K, Zakarian A, Speiser DE, Mak TW, Ohashi PS. 2000. TNF receptor 1 (TNFR1) and CD95 are not required for T cell deletion after virus infection but contribute to peptideinduced deletion under limited conditions. Eur. J. Immunol. 30:683–88 Badovinac VP, Tvinnereim AR, Harty JT. 2000. Regulation of antigen-specific CD8+ T cell homeostasis by perforin and interferon-gamma. Science 290:1354–58 Matloubian M, Suresh M, Glass A, Galvan M, Chow K, Whitmire JK, Walsh CM, Clark WR, Ahmed R. 1999. A role for perforin in downregulating T-cell responses during chronic viral infection. J. Virol. 73:2527–36 Sytwu HK, Liblau RS, McDevitt HO. 1996. The roles of Fas/APO-1 (CD95) and TNF in antigen-induced programmed cell death in T cell receptor transgenic mice. Immunity 5:17–30 Van Parijs L, Peterson DA, Abbas AK. 1998. The Fas/Fas ligand pathway and Bcl-2 regulate T cell responses to model self and foreign antigens. Immunity 8:265–74 Tivol EA, Borriello F, Schweitzer AN, Lynch WP, Bluestone JA, Sharpe AH. 1995. Loss of CTLA-4 leads to massive lymphoproliferation and fatal multiorgan tissue destruction, revealing a critical negative regulatory role of CTLA-4. Immunity 3:541–47 Waterhouse P, Penninger JM, Timms E, Wakeham A, Shahinian A, Lee KP, Thompson CB, Griesser H, Mak TW. 1995. Lymphoproliferative disorders with early lethality in mice deficient in CTLA4. Science 270:985–88 Ranger AM, Oukka M, Rengarajan J, Glimcher LH. 1998. Inhibitory function of two NFAT family members in lymphoid homeostasis and Th2 development. Immunity 9:627–35 Kneitz B, Herrmann T, Yonehara S,

4 Feb 2002

11:50

AR

AR152-18.tex

AR152-18.SGM

LaTeX2e(2001/05/10)

P1: GKD

T CELL MEMORY

Annu. Rev. Immunol. 2002.20:551-579. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

104.

105.

106.

107.

108.

109.

110.

111.

Schimpl A. 1995. Normal clonal expansion but impaired Fas-mediated cell death and anergy induction in interleukin-2deficient mice. Eur. J. Immunol. 25:2572– 77 Van Parijs L, Biuckians A, Ibragimov A, Alt FW, Willerford DM, Abbas AK. 1997. Functional responses and apoptosis of CD25 (IL-2R alpha)-deficient T cells expressing a transgenic antigen receptor. J. Immunol. 158:3738–45 Tsunobuchi H, Nishimura H, Goshima F, Daikoku T, Nishiyama Y, Yoshikai Y. 2000. Memory-type CD8+ T cells protect IL-2 receptor alpha-deficient mice from systemic infection with herpes simplex virus type 2. J. Immunol. 165:4552–60 Suzuki H, Kundig TM, Furlonger C, Wakeham A, Timms E, Matsuyama T, Schmits R, Simard JJ, Ohashi PS, Griesser H, et al. 1995. Deregulated T cell activation and autoimmunity in mice lacking interleukin-2 receptor beta. Science 268:1472–76 Majeti R, Xu Z, Parslow TG, Olson JL, Daikh DI, Killeen N, Weiss A. 2000. An inactivating point mutation in the inhibitory wedge of CD45 causes lymphoproliferation and autoimmunity. Cell 103:1059–70 Lucas PJ, Kim SJ, Melby SJ, Gress RE. 2000. Disruption of T cell homeostasis in mice expressing a T cell-specific dominant negative transforming growth factor beta II receptor. J. Exp. Med. 191:1187– 96 Jevnikar AM, Grusby MJ, Glimcher LH. 1994. Prevention of nephritis in major histocompatibility complex class II-deficient MRL-lpr mice. J. Exp. Med. 179:1137–43 Koh DR, Ho A, Rahemtulla A, FungLeung WP, Griesser H, Mak TW. 1995. Murine lupus in MRL/lpr mice lacking CD4 or CD8 T cells. Eur. J. Immunol. 25:2558–62 Chambers CA, Sullivan TJ, Allison JP. 1997. Lymphoproliferation in CTLA-4deficient mice is mediated by costi-

112.

113.

114.

115.

116.

117.

118.

119.

120.

121.

575

mulation-dependent activation of CD4+ T cells. Immunity 7:885–95 Nishimura H, Honjo T. 2001. PD-1: an inhibitory immunoreceptor involved in peripheral tolerance. Trends Immunol. 22:265–68 Sadlack B, Lohler J, Schorle H, Klebb G, Haber H, Sickel E, Noelle RJ, Horak I. 1995. Generalized autoimmune disease in interleukin-2-deficient mice is triggered by an uncontrolled activation and proliferation of CD4+ T cells. Eur. J. Immunol. 25:3053–59 Chambers CA, Allison JP. 1997. Costimulation in T cell responses. Curr. Opin. Immunol. 9:396–404 Siegel RM, Chan FK, Chun HJ, Lenardo MJ. 2000. The multifaceted role of Fas signaling in immune cell homeostasis and autoimmunity. Nat. Immunol. 1:469–74 Tschopp J, Irmler M, Thome M. 1998. Inhibition of fas death signals by FLIPs. Curr. Opin. Immunol. 10:552–58 Scaffidi C, Schmitz I, Krammer PH, Peter ME. 1999. The role of c-FLIP in modulation of CD95-induced apoptosis. J. Biol. Chem. 274:1541–48 Dalton DK, Haynes L, Chu CQ, Swain SL, Wittmer S. 2000. Interferon gamma eliminates responding CD4 T cells during mycobacterial infection by inducing apoptosis of activated CD4 T cells. J. Exp. Med. 192:117–22 Zheng L, Fisher G, Miller RE, Peschon J, Lynch DH, Lenardo MJ. 1995. Induction of apoptosis in mature T cells by tumour necrosis factor. Nature 377:348–51 Strasser A. 1995. Life and death during lymphocyte development and function: evidence for two distinct killing mechanisms. Curr. Opin. Immunol. 7:228–34 Moreno MB, Memon SA, Zacharchuk CM. 1996. Apoptosis signaling pathways in normal T cells: differential activity of Bcl-2 and IL-1beta-converting enzyme family protease inhibitors on glucocorticoid- and Fas-mediated cytotoxicity. J. Immunol. 157:3845–49

4 Feb 2002

11:50

Annu. Rev. Immunol. 2002.20:551-579. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

576

AR

AR152-18.tex

SPRENT

¥

AR152-18.SGM

LaTeX2e(2001/05/10)

P1: GKD

SURH

122. Arstila TP, Casrouge A, Baron V, Even J, Kanellopoulos J, Kourilsky P. 1999. A direct estimate of the human alpha beta T cell receptor diversity. Science 286:958– 61 123. McHeyzer-Williams LJ, Panus JF, Mikszta JA, McHeyzer-Williams MG. 1999. Evolution of antigen-specific T cell receptors in vivo: preimmune and antigen-driven selection of preferred complementarity-determining region 3 (CDR3) motifs. J. Exp. Med. 189:1823– 38 124. Busch DH, Pamer EG. 1999. T cell affinity maturation by selective expansion during infection. J. Exp. Med. 189:701–10 125. Savage PA, Boniface JJ, Davis MM. 1999. A kinetic basis for T cell receptor repertoire selection during an immune response. Immunity 10:485–92 126. Rees W, Bender J, Teague TK, Kedl RM, Crawford F, Marrack P, Kappler J. 1999. An inverse relationship between T cell receptor affinity and antigen dose during CD4+ T cell responses in vivo and in vitro. Proc. Natl. Acad. Sci. USA 96:9781–86 127. Kedl RM, Rees WA, Hildeman DA, Schaefer B, Mitchell T, Kappler J, Marrack P. 2000. T cells compete for access to antigen-bearing antigen-presenting cells. J. Exp. Med. 192:1105–13 128. Yewdell JW, Bennink JR. 1999. Immunodominance in major histocompatibility complex class I-restricted T lymphocyte responses. Annu. Rev. Immunol. 17:51–88 129. Busch DH, Pilip I, Pamer EG. 1998. Evolution of a complex T cell receptor repertoire during primary and recall bacterial infection. J. Exp. Med. 188:61–70 130. Sourdive DJ, Murali-Krishna K, Altman JD, Zajac AJ, Whitmire JK, Pannetier C, Kourilsky P, Evavold B, Sette A, Ahmed R. 1998. Conserved T cell receptor repertoire in primary and memory CD8 T cell responses to an acute viral infection. J. Exp. Med. 188:71–82 131. Zheng B, Han S, Zhu Q, Goldsby R, Kelsoe G. 1996. Alternative pathways for the

132.

133.

134.

135.

136.

137.

138.

139. 140. 141.

142.

selection of antigen-specific peripheral T cells. Nature 384:263–66 Panus JF, McHeyzer-Williams LJ, McHeyzer-Williams MG. 2000. Antigenspecific T helper cell function: differential cytokine expression in primary and memory responses. J. Exp. Med. 192:1301–16 Sallusto F, Lenig D, Forster R, Lipp M, Lanzavecchia A. 1999. Two subsets of memory T lymphocytes with distinct homing potentials and effector functions. Nature 401:708–12 Rogers PR, Dubey C, Swain SL. 2000. Qualitative changes accompany memory T cell generation: faster, more effective responses at lower doses of antigen. J. Immunol. 164:2338–46 Dai Z, Konieczny BT, Lakkis FG. 2000. The dual role of IL-2 in the generation and maintenance of CD8+ memory T cells. J. Immunol. 165:3031–36 Opferman JT, Ober BT, Ashton-Rickardt PG. 1999. Linear differentiation of cytotoxic effectors into memory T lymphocytes. Science 283:1745–48 Jacob J, Baltimore D. 1999. Modelling T cell memory by genetic marking of memory T cells in vivo. Nature 399:593–97 Saparov A, Wagner FH, Zheng R, Oliver JR, Maeda H, Hockett RD, Weaver CT. 1999. Interleukin-2 expression by a subpopulation of primary T cells is linked to enhanced memory/effector function. Immunity 11:271–80 Sprent J. 1994. T and B memory cells. Cell 76:315–22 Sprent J, Tough DF. 2001. T cell death and memory. Science 293:245–48 Moskophidis D, Lechner F, Pircher H, Zinkernagel R. 1993. Virus persistence in acutely infected immunocompetent mice by exhaustion of antiviral cytotoxic effector T cells. Nature 362:758–61 Garcia S, DiSanto J, Stockinger B. 1999. Following the development of a CD4 T cell response in vivo: from activation to memory formation. Immunity 11:163– 71

4 Feb 2002

11:50

AR

AR152-18.tex

AR152-18.SGM

LaTeX2e(2001/05/10)

P1: GKD

Annu. Rev. Immunol. 2002.20:551-579. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

T CELL MEMORY 143. Sad S, Kagi D, Mosmann TR. 1996. Perforin and Fas killing by CD8+ T cells limits their cytokine synthesis and proliferation. J. Exp. Med. 184:1543–47 144. Hu H, Huston G, Duso D, Lepak N, Roman E, Swain SL. 2001. CD4+ T cell effectors can become memory cells with high efficiency and without further division. Nat. Immunol. 2:705–10 145. Surh CD, Sprent J. 2000. Homeostatic T cell proliferation. How far can T cells be activated to self-ligands? J. Exp. Med. 192:F9–14 146. Tripp RA, Hou S, Doherty PC. 1995. Temporal loss of the activated L-selectinlow phenotype for virus-specific CD8+ T memory cells. J. Immunol. 154:5870– 75 147. Walker PR, Ohteki T, Lopez JA, MacDonald HR, Maryanski JL. 1995. Distinct phenotypes of antigen-selected CD8 T cells emerge at different stages of an in vivo immune response. J. Immunol. 155:3443–52 148. Zimmerman C, Brduscha-Riem K, Blaser C, Zinkernagel RM, Pircher H. 1996. Visualization, characterization, and turnover of CD8+ memory T cells in virusinfected hosts. J. Exp. Med. 183:1367–75 149. Kedl RM, Mescher MF. 1998. Qualitative differences between naive and memory T cells make a major contribution to the more rapid and efficient memory CD8+ T cell response. J. Immunol. 161:674– 83 150. Pihlgren M, Dubois PM, Tomkowiak M, Sjogren T, Marvel J. 1996. Resting memory CD8+ T cells are hyperreactive to antigenic challenge in vitro. J. Exp. Med. 184:2141–51 151. Bunce C, Bell EB. 1997. CD45RC isoforms define two types of CD4 memory T cells, one of which depends on persisting antigen. J. Exp. Med. 185:767–76 152. Cho BK, Wang C, Sugawa S, Eisen HN, Chen J. 1999. Functional differences between memory and naive CD8 T cells. Proc. Natl. Acad. Sci. USA 96:2976–81

577

153. Yawalkar N, Hunger RE, Pichler WJ, Braathen LR, Brand CU. 2000. Human afferent lymph from normal skin contains an increased number of mainly memory/effector CD4+ T cells expressing activation, adhesion and co-stimulatory molecules. Eur. J. Immunol. 30:491– 97 154. Hogan RJ, Usherwood EJ, Zhong W, Roberts AA, Dutton RW, Harmsen AG, Woodland DL. 2001. Activated antigenspecific CD8+ T cells persist in the lungs following recovery from respiratory virus infections. J. Immunol. 166:1813–22 155. Hogan RJ, Zhong W, Usherwood EJ, Cookenham T, Roberts AD, Woodland DL. 2001. Protection from respiratory virus infections can be mediated by antigen-specific CD4+ T cells that persist in the lungs. J. Exp. Med. 193:981– 86 156. Baron JL, Madri JA, Ruddle NH, Hashim G, Janeway CA Jr. 1993. Surface expression of alpha 4 integrin by CD4 T cells is required for their entry into brain parenchyma. J. Exp. Med. 177:57–68 157. Veiga-Fernandes H, Walter U, Bourgeois C, McLean A, Rocha B. 2000. Response of naive and memory CD8+ T cells to antigen stimulation in vivo. Nat. Immunol. 1:47–53 158. Bachmann MF, Barner M, Viola A, Kopf M. 1999. Distinct kinetics of cytokine production and cytolysis in effector and memory T cells after viral infection. Eur. J. Immunol. 29:291–99 159. Zimmermann C, Prevost-Blondel A, Blaser C, Pircher H. 1999. Kinetics of the response of naive and memory CD8 T cells to antigen: similarities and differences. Eur. J. Immunol. 29:284–90 160. Hou S, Hyland L, Ryan KW, Portner A, Doherty PC. 1994. Virus-specific CD8+ T cell memory determined by clonal burst size. Nature 369:652–54 161. Lau LL, Jamieson BD, Somasundaram R, Ahmed R. 1994. Cytotoxic T cell memory without antigen. Nature 369:648–52

4 Feb 2002

11:50

Annu. Rev. Immunol. 2002.20:551-579. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

578

AR

AR152-18.tex

SPRENT

¥

AR152-18.SGM

LaTeX2e(2001/05/10)

P1: GKD

SURH

162. Mullbacher A. 1994. The long-term maintenance of cytotoxic T cell memory does not require persistence of antigen. J. Exp. Med. 179:317–21 163. Markiewicz MA, Girao C, Opferman JT, Sun J, Hu Q, Agulnik AA, Bishop CE, Thompson CB, Ashton-Rickardt PG. 1998. Long-term T cell memory requires the surface expression of selfpeptide/major histocompatibility complex molecules. Proc. Natl. Acad. Sci. USA 95:3065–70 164. Beverley PCL. 1990. Is T cell memory maintained by cross-reactive stimulation? Immunol. Today 11:203–5 165. Murali-Krishna K, Lau LL, Sambhara S, Lemonnier F, Altman J, Ahmed R. 1999. Persistence of memory CD8 T cells in MHC class I-deficient mice. Science 286:1377–81 166. Swain SL, Hu H, Huston G. 1999. Class II-independent generation of CD4 memory T cells from effectors. Science 286:1381–83 167. Tough DF, Sprent J. 1994. Turnover of naive- and memory-phenotype T cells. J. Exp. Med. 179:1127–35 168. Tough DF, Borrow P, Sprent J. 1996. Induction of bystander T cell proliferation by viruses and type I interferon in vivo. Science 272:1947–50 169. Tough DF, Sun S, Sprent J. 1997. T cell stimulation in vivo by lipopolysaccharide (LPS). J. Exp. Med. 185:2089–94 170. Sprent J, Zhang X, Sun S, Tough D. 2000. T cell proliferation in vivo and the role of cytokines. Philos. Trans. R. Soc. London B Ser. 355:317–22 171. Tough DF, Zhang X, Sprent J. 2001. An IFN-γ -dependent pathway controls stimulation of memory-phenotype CD8+ T cell turnover in vivo by IL-12, IL-18 and IFN-g. J. Immunol. 166:6007–11 172. Zhang X, Sun S, Hwang I, Tough DF, Sprent J. 1998. Potent and selective stimulation of memory-phenotype CD8+ T cells in vivo by IL-15. Immunity 8:591– 99

173. Judge AD, Zhang X, Fujii H, Surh CD, Sprent J. 2002. IL-15 controls both proliferation and survival of a subset of memory-phenotype CD8+ T cells. Submitted 174. Ku CC, Murakami M, Sakamoto A, Kappler J, Marrack P. 2000. Control of homeostasis of CD8+ memory T cells by opposing cytokines. Science 288:675– 78 175. Kennedy MK, Glaccum M, Brown SN, Butz EA, Viney JL, Embers M, Matsuki N, Charrier K, Sedger L, Willis CR, Brasel K, Morrissey PJ, Stocking K, Schuh JC, Joyce S, Peschon JJ. 2000. Reversible defects in natural killer and memory CD8 T cell lineages in interleukin 15-deficient mice. J. Exp. Med. 191:771–80 176. Nishimura H, Yajima T, Naiki Y, Tsunobuchi H, Umemura M, Itano K, Matsuguchi T, Suzuki M, Ohashi PS, Yoshikai Y. 2000. Differential roles of interleukin 15 mRNA isoforms generated by alternative splicing in immune responses in vivo. J. Exp. Med. 191:157–70 177. Marks-Konczalik J, Dubois S, Losi JM, Sabzevari H, Yamada N, Feigenbaum L, Waldmann TA, Tagaya Y. 2000. IL-2induced activation-induced cell death is inhibited in IL-15 transgenic mice. Proc. Natl. Acad. Sci. USA 97:11,445–50 178. Khan IA, Casciotti L. 1999. IL-15 prolongs the duration of CD8+ T cellmediated immunity in mice infected with a vaccine strain of Toxoplasma gondii. J. Immunol. 163:4503–9 179. Lodolce JP, Boone DL, Chai S, Swain RE, Dassopoulos T, Trettin S, Ma A. 1998. IL-15 receptor maintains lymphoid homeostasis by supporting lymphocyte homing and proliferation. Immunity 9:669–76 180. Huang LR, Chen FL, Chen YT, Lin YM, Kung JT. 2000. Potent induction of longterm CD8+ T cell memory by shortterm IL-4 exposure during T cell receptor stimulation. Proc. Natl. Acad. Sci. USA 97:3406–11

4 Feb 2002

11:50

AR

AR152-18.tex

AR152-18.SGM

LaTeX2e(2001/05/10)

P1: GKD

Annu. Rev. Immunol. 2002.20:551-579. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

T CELL MEMORY 181. Lantz O, Grandjean I, Matzinger P, Di Santo JP. 2000. Gamma chain required for naive CD4+ T cell survival but not for antigen proliferation. Nat. Immunol. 1:54–58 182. Scollay R, Sprent J, eds. 1997. Lymphocyte homeostasis. Semin. Immunol. 9:329–404 183. Tanchot C, Fernandes HV, Rocha B. 2000. The organization of mature T-cell pools. Philos. Trans. R. Soc. London B Ser. 355:323–28 184. Akbar AN, Salmon M, Savill J, Janossy G. 1993. A possible role for bcl-2 in regulating T-cell memory—a balancing act between cell death and survival. Immunol. Today 14:526–32 185. Zhang X, Fujii H, Kishimoto H, LeRoy E, Surh CD, Sprent J. 2002. Aging leads to disturbed homeostasis of memoryphenotype CD8+ cells. Submitted 186. Pilling D, Akbar AN, Girdlestone J, Orteu CH, Borthwick NJ, Amft N, ScheelToellner D, Buckley CD, Salmon M. 1999. Interferon-beta mediates stromal cell rescue of T cells from apoptosis. Eur. J. Immunol. 29:1041–50 187. Grayson JM, Zagac AJ, Altman JD, Ahmed R. 2000. Cutting edge: increased expression of Bcl-2 in antigen-specific memory CD8+ T cells. J. Immunol. 164: 3950–54 188. Buckley AF, Kuo CT, Leiden JM. 2001. Transcription factor LKLF is sufficient

189.

190.

191.

192.

193.

579

to program T cell quiescence via a cMyc-dependent pathway. Nat. Immunol. 2:698–704 Schober SL, Kuo CT, Schluns KS, Lefrancois L, Leiden JM, Jameson SC. 1999. Expression of the transcription factor lung Kruppel-like factor is regulated by cytokines and correlates with survival of memory T cells in vitro and in vivo. J. Immunol. 163:3662–67 Mertsching E, Burdet C, Ceredig R. 1995. IL-7 transgenic mice: analysis of the role of IL-7 in the differentiation of thymocytes in vivo and in vitro. Int. Immunol. 7:401–14 Suzuki H, Zhou YW, Kato M, Mak TW, Nakashima I. 1999. Normal regulatory alpha/beta T cells effectively eliminate abnormally activated T cells lacking the interleukin 2 receptor beta in vivo. J. Exp. Med. 190:1561–72 Annacker O, Burlen-Defranoux O, Pimenta-Araujo R, Cumano A, Bandeira A. 2000. Regulatory CD4 T cells control the size of the peripheral activated/memory CD4 T cell compartment. J. Immunol. 164:3573–80 Zarozinski CC, McNally JM, Lohman BL, Daniels KA, Welsh RM. 2000. Bystander sensitization to activationinduced cell death as a mechanism of virus-induced immune suppression. J. Virol. 74:3650–58

P1: FRK

February 1, 2002

12:47

Annual Reviews

AR152-FM

Annual Review of Immunology Volume 20, 2002

Annu. Rev. Immunol. 2002.20:551-579. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

CONTENTS FRONTISPIECE, Charles A. Janeway, Jr. A TRIP THROUGH MY LIFE WITH AN IMMUNOLOGICAL THEME, Charles A. Janeway, Jr.

xiv 1

THE B7 FAMILY OF LIGANDS AND ITS RECEPTORS: NEW PATHWAYS FOR COSTIMULATION AND INHIBITION OF IMMUNE RESPONSES, Beatriz M. Carreno and Mary Collins

29

MAP KINASES IN THE IMMUNE RESPONSE, Chen Dong, Roger J. Davis, and Richard A. Flavell

PROSPECTS FOR VACCINE PROTECTION AGAINST HIV-1 INFECTION AND AIDS, Norman L. Letvin, Dan H. Barouch, and David C. Montefiori T CELL RESPONSE IN EXPERIMENTAL AUTOIMMUNE ENCEPHALOMYELITIS (EAE): ROLE OF SELF AND CROSS-REACTIVE ANTIGENS IN SHAPING, TUNING, AND REGULATING THE AUTOPATHOGENIC T CELL REPERTOIRE, Vijay K. Kuchroo, Ana C. Anderson, Hanspeter Waldner, Markus Munder, Estelle Bettelli, and Lindsay B. Nicholson

55 73

101

NEUROENDOCRINE REGULATION OF IMMUNITY, Jeanette I. Webster, Leonardo Tonelli, and Esther M. Sternberg

125

MOLECULAR MECHANISM OF CLASS SWITCH RECOMBINATION: LINKAGE WITH SOMATIC HYPERMUTATION, Tasuku Honjo, Kazuo Kinoshita, and Masamichi Muramatsu

165

INNATE IMMUNE RECOGNITION, Charles A. Janeway, Jr. and Ruslan Medzhitov

KIR: DIVERSE, RAPIDLY EVOLVING RECEPTORS OF INNATE AND ADAPTIVE IMMUNITY, Carlos Vilches and Peter Parham ORIGINS AND FUNCTIONS OF B-1 CELLS WITH NOTES ON THE ROLE OF CD5, Robert Berland and Henry H. Wortis E PROTEIN FUNCTION IN LYMPHOCYTE DEVELOPMENT, Melanie W. Quong, William J. Romanow, and Cornelis Murre

197 217 253 301

LYMPHOCYTE-MEDIATED CYTOTOXICITY, John H. Russell and Timothy J. Ley

SIGNAL TRANSDUCTION MEDIATED BY THE T CELL ANTIGEN x

323

P1: FRK

Annu. Rev. Immunol. 2002.20:551-579. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

February 1, 2002

12:47

Annual Reviews

AR152-FM

CONTENTS

xi

RECEPTOR: THE ROLE OF ADAPTER PROTEINS, Lawrence E. Samelson INTERACTION OF HEAT SHOCK PROTEINS WITH PEPTIDES AND ANTIGEN PRESENTING CELLS: CHAPERONING OF THE INNATE AND ADAPTIVE IMMUNE RESPONSES, Pramod Srivastava CHROMATIN STRUCTURE AND GENE REGULATION IN THE IMMUNE SYSTEM, Stephen T. Smale and Amanda G. Fisher PRODUCING NATURE’S GENE-CHIPS: THE GENERATION OF PEPTIDES FOR DISPLAY BY MHC CLASS I MOLECULES, Nilabh Shastri, Susan

371

Schwab, and Thomas Serwold

395 427

463

THE IMMUNOLOGY OF MUCOSAL MODELS OF INFLAMMATION, Warren Strober, Ivan J. Fuss, and Richard S. Blumberg T CELL MEMORY, Jonathan Sprent and Charles D. Surh

GENETIC DISSECTION OF IMMUNITY TO MYCOBACTERIA: THE HUMAN MODEL, Jean-Laurent Casanova and Laurent Abel ANTIGEN PRESENTATION AND T CELL STIMULATION BY DENDRITIC CELLS, Pierre Guermonprez, Jenny Valladeau, Laurence Zitvogel, Clotilde Th´ery, and Sebastian Amigorena

495 551 581

621

NEGATIVE REGULATION OF IMMUNORECEPTOR SIGNALING, Andr´e Veillette, Sylvain Latour, and Dominique Davidson

669

CPG MOTIFS IN BACTERIAL DNA AND THEIR IMMUNE EFFECTS, Arthur M. Krieg

PROTEIN KINASE Cθ

709 IN

T CELL ACTIVATION, Noah Isakov and Amnon

Altman

RANK-L AND RANK: T CELLS, BONE LOSS, AND MAMMALIAN EVOLUTION, Lars E. Theill, William J. Boyle, and Josef M. Penninger PHAGOCYTOSIS OF MICROBES: COMPLEXITY IN ACTION, David M. Underhill and Adrian Ozinsky

761 795 825

STRUCTURE AND FUNCTION OF NATURAL KILLER CELL RECEPTORS: MULTIPLE MOLECULAR SOLUTIONS TO SELF, NONSELF DISCRIMINATION, Kannan Natarajan, Nazzareno Dimasi, Jian Wang, Roy A. Mariuzza, and David H. Margulies

853

INDEXES Subject Index Cumulative Index of Contributing Authors, Volumes 1–20 Cumulative Index of Chapter Titles, Volumes 1–20

ERRATA An online log of corrections to Annual Review of Immunology chapters (if any, 1997 to the present) may be found at http://immunol.annualreviews.org/

887 915 925

8 Feb 2002

10:4

AR

AR152-19.tex

AR152-19.SGM

LaTeX2e(2001/05/10) P1: GJC 10.1146/annurev.immunol.20.081501.125851

Annu. Rev. Immunol. 2002. 20:581–620 DOI: 10.1146/annurev.immunol.20.081501.125851 c 2002 by Annual Reviews. All rights reserved Copyright °

GENETIC DISSECTION OF IMMUNITY TO MYCOBACTERIA: The Human Model Annu. Rev. Immunol. 2002.20:581-620. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

Jean-Laurent Casanova and Laurent Abel Laboratory of Human Genetics of Infectious Diseases, Universit´e Ren´e Descartes-INSERM U550, Necker Medical School, 156 rue de Vaugirard, 75015 Paris, France, European Union; e-mail: [email protected], [email protected]

Key Words Bacille Calmette-Gu´erin, environmental mycobacteria, leprosy, tuberculosis, Mendelian inheritance, complex inheritance ■ Abstract Humans are exposed to a variety of environmental mycobacteria (EM), and most children are inoculated with live Bacille Calmette-Gu´erin (BCG) vaccine. In addition, most of the world’s population is occasionally exposed to human-borne mycobacterial species, which are less abundant but more virulent. Although rarely pathogenic, mildly virulent mycobacteria, including BCG and most EM, may cause a variety of clinical diseases. Mycobacterium tuberculosis, M. leprae, and EM M. ulcerans are more virulent, causing tuberculosis, leprosy, and Buruli ulcer, respectively. Remarkably, only a minority of individuals develop clinical disease, even if infected with virulent mycobacteria. The interindividual variability of clinical outcome is thought to result in part from variability in the human genes that control host defense. In this well-defined microbiological and clinical context, the principles of mouse immunology and the methods of human genetics can be combined to facilitate the genetic dissection of immunity to mycobacteria. The natural infections are unique to the human model, not being found in any of the animal models of experimental infection. We review current genetic knowledge concerning the simple and complex inheritance of predisposition to mycobacterial diseases in humans. Rare patients with Mendelian disorders have been found to be vulnerable to BCG, a few EM, and M. tuberculosis. Most cases of presumed Mendelian susceptibility to these and other mycobacterial species remain unexplained. In the general population leprosy and tuberculosis have been shown to be associated with certain human genetic polymorphisms and linked to certain chromosomal regions. The causal vulnerability genes themselves have yet to be identified and their pathogenic alleles immunologically validated. The studies carried out to date have been fruitful, initiating the genetic dissection of protective immunity against a variety of mycobacterial species in natural conditions of infection. The human model has potential uses beyond the study of mycobacterial infections and may well become a model of choice for the investigation of immunity to infectious agents.

0732-0582/02/0407-0581$14.00

581

2 Feb 2002

11:3

582

AR

AR152-19.tex

CASANOVA

¥

AR152-19.SGM

LaTeX2e(2001/05/10)

P1: GJC

ABEL

Annu. Rev. Immunol. 2002.20:581-620. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

INTRODUCTION The genus Mycobacterium, together with Corynebacterium and Nocardia, forms a monophyletic taxon within the family of Actinomycetes (1). These gram-positive bacteria have very waxy cell walls owing to the presence of mycolic acids. This renders them acid- and alcohol-fast, a feature that specifically distinguishes them from other bacteria. Unique biochemical pathways differentiate Mycobacteria from the related Corynebacterium and Nocardia genera. Extensive biochemical and genetic studies have made mycobacteria one of the groups of microorganisms about which we know the most (2, 3). The genus Mycobacterium is highly diverse, with 85 different species identified since the identification of M. leprae in 1873 (4). In addition there are almost certainly many more species that remain to be discovered. Based on sequencing of the 16S RNA (5), RNA polymerase (RpoB) (6), or hsp 65 (7) genes, species can be rapidly identified and phylogenetic trees established, illustrating the diversity of the mycobacterial world. This genotypic classification is consistent with the phenotypic classification of cultivable strains into rapidly growing and slowly growing species in vitro (2). There are also a number of Bacille Calmette-Gu´erin (BCG) vaccine substrains, derived from an attenuated M. bovis strain obtained in 1921 (8). Finally, the individual Mycobacterium species, such as M. tuberculosis, display significant diversity (9). The vast majority of mycobacterial species are environmental free-living saprophytes (10). Mycobacteria have adapted to various environmental conditions and grow in the soil (11) and water (12) of various regions of the world. They can live in a variety of natural waters, including fresh- and saltwater, and treated water, including swimming pools and drinking water, from which they are readily spread via aerosols. However, little is known about the metabolic requirements of these bacteria in their natural niches. For example, in drinking water distribution systems, M. avium is preferentially recovered from water samples, whereas the closely related M. intracellulare is recovered from biofilms (13). A few mycobacterial species, such as M. tuberculosis and M. bovis, first identified in infected humans and cattle, respectively, but both capable of infecting other animal species, have never been identified in the environment (14–16). This suggests that these organisms are obligate parasites of humans or animals. However, definitive conclusions cannot be drawn, as there may be many unexplored environmental niches. For example, M. ulcerans was only recently identified in the environment (17, 18). Along these lines, M. leprae was thought to be strictly human-tropic until natural infection of a rare animal, the nine-banded armadillo, was documented (19, 20). Caution is therefore required in observations concerning the natural reservoirs and hosts of mycobacteria and their resulting classification. All humans are exposed to water- and air-borne environmental mycobacteria (EM), which frequently come into contact with the skin and mucous membranes (principally the digestive and respiratory epithelia). In addition, the vast majority (85%) of children worldwide are inoculated intradermally or subcutaneously with live BCG vaccine. A large proportion of humans, mostly in the poorest countries

2 Feb 2002

11:3

AR

AR152-19.tex

AR152-19.SGM

LaTeX2e(2001/05/10)

P1: GJC

Annu. Rev. Immunol. 2002.20:581-620. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

HUMAN GENETICS OF MYCOBACTERIA

583

or regions of the world, are also exposed to primarily air-borne human-tropic mycobacteria such as M. tuberculosis and M. leprae. Although some EM can infect animals (21) and certain mycobacterial species appear to be animal-tropic (e.g., M. bovis) (16), improvements in hygiene have reduced the level of human exposure to these species via contact with infected animals. Mycobacteria do not seem to engage in commensalism, as they have never been cultured from the skin or mucous membranes of healthy individuals (whereas they can be cultured from the tissues of infected patients). It is unclear whether humans and mycobacteria derive indirect benefit from each other via the environment. There seems to be little if any direct mutual benefit between Homo sapiens, a recently evolved vertebrate species (22, 23), and any species of the ancient Mycobacterium genus (24, 25). Rather, mycobacteria cause a broad epidemiological, clinical, and pathological spectrum of diseases in humans. M. leprae is the causal agent of leprosy, 700,000 new cases of which are reported annually worldwide (26). M. leprae resides within Schwann cells and macrophages. Although rarely life-threatening, leprosy is a chronic granulomatous disease of the skin and peripheral nerves and presents either as polar leprosy (lepromatous/multibacillary or tuberculoid/paucibacillary) or as one of several intermediate forms (19, 20). M. tuberculosis and related species (e.g., M. bovis) cause tuberculosis (14–16). About 8 million new cases of tuberculosis are reported annually worldwide, resulting in almost 1.9 million deaths (27). M. tuberculosis survives within macrophages, often despite the formation of surrounding granulomas. Tuberculosis is primarily a pulmonary disease, but organs other than the lungs may be affected. Buruli ulcer, caused by environmental M. ulcerans, is the third most common mycobacterial disease (18). The global incidence of this disease is not known, but the number of cases reported has increased in recent years (28). Unlike other mycobacteria, M. ulcerans grows extracellularly, and its pathogenicity seems to result largely from the secretion of a toxin. This chronic disease causes painless, expanding skin ulcers. Many other EM (e.g., M. avium) may cause occasional cases of localized or disseminated clinical disease, with each species being potentially pleiotropic yet generally showing a particular pattern of tissue-tropism (10). Similarly, BCG vaccines may cause local or disseminated disease (29). It is difficult to distinguish between exposure to and infection with mycobacteria other than BCG (which is inoculated) because individual host exposure cannot be strictly ascertained (30) (Figure 1). Exposure to most mycobacteria probably results only rarely in infection. The frequency of infection is itself probably underestimated because it is often based on the detection of immunological phenotypes, which reflect an adaptive memory immune response. Innate immunity may suffice to control the infection, and possible poor memory responses may make it difficult to identify the phenotype of interest. It is clear that only a minority of the individuals infected go on to develop clinical disease. This holds true not only for mildly virulent EM and BCG but also for the more virulent species M. tuberculosis (31) and M. leprae (20), which cause clinical disease in less than 10% of infected individuals. It may also be true for M. ulcerans (17). Thus, vulnerability to mycobacteria is the

2 Feb 2002

11:3

Annu. Rev. Immunol. 2002.20:581-620. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

584

AR

AR152-19.tex

CASANOVA

¥

AR152-19.SGM

LaTeX2e(2001/05/10)

P1: GJC

ABEL

Figure 1 The various steps in the interaction between humans and mycobacteria. Exposure to mycobacteria does not always result in infection. Whether or not an established infection further develops depends on innate immunity, alone or in conjunction with adaptive immunity. Immunological and clinical phenotypes may be detectable once mycobacterial infection is established and adaptive immunity to mycobacteria is involved. Each of the three steps in this process is under host and environmental control. Host factors may be genetic (e.g., mutation in a gene involved in immunity to mycobacteria) or nongenetic (e.g., skin lesion) and may have an impact at each stage of the interaction. Environmental factors may be mycobacterial (e.g., virulence factors) or related to the mode of exposure (e.g., direct inoculation) and may have an impact at each stage of the interaction.

exception rather than the rule in humans. A combination of environmental microbial and nonmicrobial factors and host genetic and nongenetic factors determines the outcome of exposure and infection. Whatever the relative contributions of these factors, the occurrence of clinical disease implies that host defense to mycobacteria has failed. Immunity to mycobacteria has been extensively investigated in several animal models (32). The greatest strides forward have been made in the mouse model, particularly since the infection of a series of knockout mice with mycobacteria (mostly BCG and M. tuberculosis and more recently M. avium). These remarkable studies (33–35, 35a) established the relative contributions of various cell subsets, such as α/β and γ /δ T cells and CD4 and CD8 α/β T cells, to the destruction of mycobacteria by macrophages. They also identified a number of key molecules involved in recognition (e.g., Tlr2), regulation (e.g., IL-12, IFNγ , TNFα), and effector (e.g., NO) processes. An inherent limitation of these genotype-to-phenotype

2 Feb 2002

11:3

AR

AR152-19.tex

AR152-19.SGM

LaTeX2e(2001/05/10)

P1: GJC

Annu. Rev. Immunol. 2002.20:581-620. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

HUMAN GENETICS OF MYCOBACTERIA

585

studies is that they require the definition of candidate genes. It has been shown, in a study using a phenotype-to-genotype approach, that the natural vulnerability of inbred mouse strains to mycobacteria is controlled by a recessive allele of the autosomal Bcg locus (36). A positional cloning approach has been used to identify a natural mutation in Nramp1/Bcg, which encodes a phagosomal membrane protein (37). This remarkable series of investigations has opened up new avenues of research into mycobacterial immunity (38, 39). Novel Mendelian mycobacterial vulnerability genes are unlikely to be identified in inbred strains of mice, and further progress will probably be achieved by investigations of complex vulnerability to mycobacteria (40). Despite its considerable impact in mycobacterial research, the mouse model suffers from the inherent limitation of being an experimental model of infection. The mycobacterial species with which the mice are inoculated are not natural mouse pathogens. M. microti is the natural agent of rodent tuberculosis, but mice are generally infected with M. tuberculosis, M. bovis BCG, or M. avium in this experimental model. In addition, the modes of infection used (e.g., the intravenous inoculation of mice kept in an artificial environment with large amounts of a laboratory mycobacterial strain) differ markedly from natural modes of infection (e.g., exposure of wild animals to low densities of natural, air-borne mycobacteria). Finally, laboratory mice form a small and poorly diverse group of inbred strains, which are less resistant and healthy than outbred wild mice. In contrast, humans form an expanding, outbred population in which most mycobacterial infections are natural, with BCG infection paradoxically serving as an “experimental” control. In addition, as human medicine is more developed than veterinary medicine, a comprehensive and up-to-date description of the phenotypic traits associated with each human mycobacterial disease is available. Finally, the technical difficulties that previously hampered biological research in humans have been largely overcome by recent major advances in human genetics (Figure 2). Considerable efforts are now focused on the genetic dissection of human protective immunity to mycobacteria in natural conditions of infection. We review the current state of knowledge in this novel and rapidly expanding field.

MILDLY VIRULENT MYCOBACTERIA: BCG AND MOST ENVIRONMENTAL MYCOBACTERIA Epidemiological Studies Environmental M. ulcerans is not discussed in this section because it is clearly more virulent than other known EM. Although mildly virulent, BCG and EM other than M. ulcerans may cause clinical diseases in humans. BCG vaccination may cause a variety of infectious adverse effects, from local adenitis (BCG-itis) to disseminated disease (BCG-osis) (29). The lung is the organ most frequently damaged, but clinical EM disease displays considerable diversity, reflecting the diversity of EM species and predisposing factors (10, 41, 42). For example,

2 Feb 2002

11:3

Annu. Rev. Immunol. 2002.20:581-620. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

586

AR

AR152-19.tex

CASANOVA

¥

AR152-19.SGM

LaTeX2e(2001/05/10)

P1: GJC

ABEL

Figure 2 Methods and strategies for identifying human mycobacterial susceptibility genes. The molecular basis of rare Mendelian predisposition to mycobacterial disease may be investigated using several strategies. Linkage analysis is usually the first step in the positional cloning approach, although the identification of visible cytogenetic abnormalities may be helpful. The candidate gene approach (“by hypothesis”) involves the prior selection of genes (generally based on studies of animal models in vivo or human cells in vitro, or comparison with other human inherited disorders with a related clinical phenotype), which are then tested by functional assays and/or mutation detection. Another potentially fruitful strategy is based on studying the differential expression of genes in tissues from affected and healthy individuals. To determine the molecular basis of complex predisposition to common mycobacterial diseases, linkage studies (which may be model-based or model-free) search for a chromosomal region that segregates nonrandomly with the infectious disease-related phenotype of interest, within a number of families. The role of polymorphisms within candidate genes identified “by experiment” (i.e., located within this candidate region) is tested in association studies (which may be population-based or family-based). Candidate genes may also be selected by hypothesis (as in Mendelian investigations) and tested by association studies. Statistical evidence for an association should be validated by functional studies aimed at determining the impact of the polymorphism studied on gene function and, potentially, on the mycobacterial infection-related phenotype of interest.

M. marinum is associated with swimming pool skin granulomas, whereas M. malmoense is associated with pneumonitis in Northern Europe. The clinical spectrum of local disease associated with a single species, such as M. avium, may extend from benign cervical adenitis in childhood to more severe pulmonary disease in adulthood. In addition, not only well-known M. avium, but also much less-virulent

2 Feb 2002

11:3

AR

AR152-19.tex

AR152-19.SGM

LaTeX2e(2001/05/10)

P1: GJC

Annu. Rev. Immunol. 2002.20:581-620. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

HUMAN GENETICS OF MYCOBACTERIA

587

species such as M. smegmatis and M. peregrinum may cause disseminated disease (43, 44). BCG and EM infections are generally considered to be rare but may be frequent in certain epidemiological contexts. Skin lesions and nosocomial inoculation favor the development of these infections (45). They are also favored by various acquired immunodeficiencies, including AIDS, hairy cell leukemia, and bone marrow and organ transplantation (10, 41, 42, 46). Human inherited disorders may favor the development of BCG/EM disease as well. Cystic fibrosis, the most frequent Mendelian disorder in Caucasian populations, results in a paucity of water in mucus secretions. Despite an apparently normal immune function, these patients suffer from chronic respiratory EM infections (47). In other patients disseminated BCG/EM disease almost invariably attests to an underlying Mendelian immune disorder, which may fall into one of two groups: classical primary immunodeficiency (PID) and Mendelian susceptibility to mycobacterial disease (MSMD) (Table 1) (46). Purely local disease (e.g., EM pneumonitis, BCG-itis) may also be favored by genetic factors, which may turn out to be simple or complex, but no epidemiological or molecular evidence is yet available either way (42).

Primary Immunodeficiencies SEVERE COMBINED IMMUNODEFICIENCY PIDs constitute a group of more than 100 Mendelian disorders (48–50). Patients with these disorders are generally susceptible to infection with a variety of viruses, bacteria, fungi, and protozoans. The genetic defect responsible predisposes the patient to severe BCG or EM disease in only a few of these disorders (46) (Table 1). Children with severe combined immunodeficiency (SCID) lack autologous T cells and are highly vulnerable to BCG (51–58), irrespective of the presence of B and/or NK cells and the underlying genetic defect. Only two SCID patients with EM disease have been reported, one with M. avium (59) and the other with M. marinum (60) disease. The small number of cases reported may be due to the low level of exposure to EM and of EM infection in children with SCID, most of whom die before their first birthday if they do not receive a bone marrow transplant. Alternatively, innate immunity may be sufficient to control EM, at least partially. Two thirds of SCID children inoculated (and hence infected) with BCG do not develop disseminated disease (58). Nevertheless, these cases demonstrate that human T cells are crucial for protective immunity against poorly virulent mycobacterial species such as BCG and some species of EM. HYPER-IgE SYNDROME Hyper-IgE syndrome (HIES) is a rare systemic autosomaldominant disorder combining a susceptibility to bacterial and fungal infections, eczema, and high serum IgE levels (48–50). The cellular defect responsible for the disease and its molecular basis have remained elusive. Disseminated BCG disease has been described in one child (61), and pneumonia caused by M. intracellulare has been reported in one adult patient (62). In some patients with HIES low levels of IL-12 and IFNγ production by blood cells may account for the occurrence

14 Feb 2002

8:46

588

AR

AR152-19.tex

CASANOVA

¥

AR152-19.SGM

LaTeX2e(2001/05/10)

P1: GJC

ABEL

Annu. Rev. Immunol. 2002.20:581-620. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

TABLE 1 Mendelian immune disorders predisposing the patient to mycobacterial diseasea Condition Clinical form and molecular basis Mycobacterium species

References

SCID

HIES

T(−)B(−)NK(−):

BCG

(51–57)

Reticular dysgenesis, ADA deficiency T(−)B(−)NK(+): RAG-1/RAG-2, Artemis defects T(−)B(+)NK(−): γ -chain, JAK-3 defects T(−)B(+)NK(+): IL-7Rα, CD45 defects

M. avium, M. marinum

(59, 60)

M. tuberculosis

(131)

BCG

(61)

M. intracellulare

(62)

XR-CGD:

BCG

(56, 64–69)

gp91-NADPH oxidase AR-CGD: p22, p47, p67-NADPH oxidase

M. avium, M. flavescens, (70–73, 133) M. fortuitum, Mycobacterium spp. M. tuberculosis (133, 134)

XR-EDA-ID: NEMO defect XR-OL-EDA-ID: NEMO defect Cleft lip/palate-EDA-ID: Not identified

M. avium, M. kansasii, Mycobacterium spp. M. chelonae M. tuberculosis

(74–79)

HIGM

XR-HIGM: CD154

M. tuberculosis, M. bovis

(132)

MSMD

Response to IFNγ abolished:

BCG

(44, 79, 90, 95, 97, 99a, 99b, 101)

CGD

EDA-ID

Not identified

c-AR-IFN-γ R1 deficiency c-AR-IFN-γ R2 deficiency

Impaired response to IFNγ : p-AR-IFN-γ R1 deficiency p-AD-IFN-γ R1 deficiency p-AR-IFN-γ R2 deficiency p-AD-STAT-1 deficiency Impaired IFNγ production: c-AR-IL-12Rβ1 deficiency c-AR-IL-12p40 deficiency a

(76, 135)

M. avium, M. kansasii, (43, 44, 90, 94, 96–102, 99b) M. szulgai, M. chelonae, M. fortuitum, M. abscessus, M. smegmatis, M. peregrinum BCG

(86, 87, 106, 107, 109a, 109b)

M. avium, M. kansasii, M.chelonae, M. abscessus, M. gorvonac, M. asiaticum M. tuberculosis

(86, 87, 90, 107, 109, 109a, 109c)

BCG

(110–113, 115)

M. avium, M. chelonae M. tuberculosis

(112–114, 116) (115)

(106)

The conditions are indicated, followed by their clinical forms and molecular defect. The Mycobacterium species isolated from patients suffering from the various conditions are indicated in three groups [BCG, EM (second line), and M. tuberculosis and related species (third line)]. References corresponding to each of the three groups of mycobacteria, for each condition or clinical form considered, are also indicated. Abbrevations: SCID, severe combined immunodeficiency; CGD, chronic granulomatous disease; HIES, hyper-IgE syndrome; EDA-ID, anhidrotic ectodermal dysplasia with immunodeficiency; OL-EDA-ID, anhidrotic ectodermal dysplasia with immunodeficiency, lymphedema, and osteopetrosis; HIGM: hyper-IgM syndrome; MSMD, Mendelian susceptibility to mycobacterial disease; AR, autosomal recessive; XR, X-linked recessive; c-AR, complete autosomal recessive; p-AR, partial autosomal recessive; p-AD: partial autosomal dominant; BCG: Bacille Calmette-Gu´erin.

2 Feb 2002

11:3

AR

AR152-19.tex

AR152-19.SGM

LaTeX2e(2001/05/10)

P1: GJC

HUMAN GENETICS OF MYCOBACTERIA

589

Annu. Rev. Immunol. 2002.20:581-620. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

of mycobacterial disease (63). However, HIES patients seem to be only mildly susceptible to BCG/EM, as many of the HIES patients reported were not only inoculated with BCG in childhood, but survived into adulthood, ruling out the possibility that they were not exposed to a variety of EM. CHRONIC GRANULOMATOUS DISEASE Patients with chronic granulomatous disease (CGD) have mutations in the genes encoding the NADPH oxidase complex, rendering phagocytes incapable of killing certain ingested microorganisms via oxygen-dependent pathways (48–50). They suffer from severe recurrent bacterial and fungal infections. Disseminated BCG-osis may occur in CGD patients (64–67), although local disease (BCG-itis) is more frequent (56, 68, 69). Disease caused by EM is less frequently reported, despite the large number of affected children and adults worldwide. Disseminated M. flavescens infection (70), M. fortuitum pneumonitis (71) and osteomyelitis (72), and M. avium pneumonitis (73) have been reported in CGD patients. These studies demonstrate that the phagocytic respiratory burst contributes to the control of weakly virulent mycobacterial species, such as BCG and a few EM. ANHIDROTIC ECTODERMAL DYSPLASIA WITH IMMUNODEFICIENCY Children with a rare multi-systemic disorder known as anhidrotic ectodermal dysplasia with immunodeficiency (EDA-ID) (48–50) are vulnerable to a variety of bacterial infections. The genetic basis for X-linked EDA-ID was recently determined. Hypomorphic mutations in the NEMO gene, which encodes an essential component of the NF-κB activation machinery, were identified in affected males. Six patients with EDA-ID presented with disseminated M. avium disease in the first three years of life (74–76). Other patients with more severe NEMO mutations and the related syndrome, EDA-ID with osteopetrosis and lymphedema (OL-EDA-ID) (75), presented with M. kansasii (77) or an unidentified EM (78) infection in the first year of life. Disseminated M. chelonae infection was diagnosed in a girl with a related syndrome consisting of cleft lip/palate and EDA-ID (79). Mycobacterial disease is thought to result from the impairment of innate and adaptive immunity, as phagocytic (TNFα-R, Tlrs) and T-cell (IL-1α-R, IL-18-R) receptors signal through NF-κB. The high frequency and severity of early-onset EM infection in EDA-ID and related syndromes, particularly in OL-EDA-ID, indicates that NEMOdependent NF-κB activation is crucial for protective immunity against EM. PRIMARY IMMUNODEFICIENCIES, BCG, AND ENVIRONMENTAL MYCOBACTERIA Unexpectedly, mycobacterial disease is only rarely associated with PIDs. However, anticipated advances in the identification of EM in clinical samples and the diagnosis of more patients with PIDs may somewhat qualify this conclusion. Only four PIDs are associated with a predisposition to BCG and EM infection, and mycobacterial disease occurs only in a minority of affected individuals (from a small percentage of patients with CGD and HIES to approximately one third of patients with SCID and EDA-ID). Remarkably, most of the known PIDs do not

2 Feb 2002

11:3

Annu. Rev. Immunol. 2002.20:581-620. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

590

AR

AR152-19.tex

CASANOVA

¥

AR152-19.SGM

LaTeX2e(2001/05/10)

P1: GJC

ABEL

predispose patients to mycobacterial infection (48–50). Complement and antibody deficiencies and more surprisingly most antigen-presenting cell and T-cell defects do not predispose patients to mycobacterial disease. Notably, these PIDs include HLA class I and HLA class II deficiencies (caused by molecular defects in trans) (46). There is conclusive evidence that NF-κB activation and T-cell development are required for immunity to BCG and EM. In contrast, antigen presentation by HLA class I or HLA class II molecules is apparently not required to ensure protective immunity to poorly virulent mycobacterial species. A number of children with severe mycobacterial disease and an ill-defined PID remain to be investigated at a molecular level, and the identification of novel mechanisms that contribute to mycobacterial immunity is therefore expected.

Inherited Disorders of the IL-12-IFNγ Axis MENDELIAN SUSCEPTIBILITY TO MYCOBACTERIAL DISEASE BCG and EM may also cause disseminated disease in otherwise healthy individuals with no classical primary immunodeficiency (80–83). These patients do not generally have associated infections, apart from salmonellosis, which affects less than half of the cases. Parental consanguinity and familial forms are frequently observed, and this syndrome was therefore named Mendelian susceptibility to mycobacterial disease (Table 1) (MIM 209950) (84). The syndrome is heterogeneous, although its clinical features seem to be restricted to a predisposition to mycobacterial infection. First, the genetic basis of the syndrome is not the same in all affected families. In most familial cases, inheritance is autosomal and recessive, but X-linked recessive inheritance seems to be involved in one family (83, 85), and autosomal dominant inheritance has been reported in several other families (86, 87, 109b,c). Second, clinical outcome differs between patients and has been found to correlate with the type of BCG granulomatous lesion present (88) and osteopontin expression in situ (88a). Children with granulomas of the lepromatous type (poorly delimited, multibacillary, with no epithelioid or giant cells) generally die of overwhelming infection, whereas patients with tuberculoid granulomas (well delimited, paucibacillary, with epithelioid and giant cells) have a favorable outcome. Positional cloning and a candidate gene approach have led to the identification of five Mendelian mycobacterial susceptibility genes that are mutated in children and adults with this syndrome: IFNGR1 and IFNGR2, encoding the two chains of the receptor for IFNγ , a pleiotropic cytokine secreted by NK and T cells; STAT1, encoding an essential transducer of IFNγ -mediated signals; IL12B, encoding the p40 subunit of IL-12, a potent IFNγ -inducing cytokine secreted by macrophages and dendritic cells; and IL12RB1, encoding the β1 chain of the receptor for IL12, expressed on NK and T cells. The type of mutation (recessive/dominant, hypomorphic/loss-of-function) also accounts for clinical heterogeneity, as the various mutations define nine disorders. All defects result in impaired IFNγ -mediated immunity. IFNγ secretion is impaired in patients with IL-12p40 and IL-12Rβ1 deficiency, whereas the response to IFNγ is impaired or abrogated in patients

2 Feb 2002

11:3

AR

AR152-19.tex

AR152-19.SGM

LaTeX2e(2001/05/10)

P1: GJC

HUMAN GENETICS OF MYCOBACTERIA

591

Annu. Rev. Immunol. 2002.20:581-620. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

with partial or complete IFNγ R1, IFNγ R2, and STAT-1 deficiencies, respectively. We review these disorders and discuss their possible immunological implications. Clinical aspects of these conditions have been recently reviewed elsewhere (89–93). COMPLETE IFNγ R1 AND IFNγ R2 DEFICIENCY Complete interferon γ receptor ligand-binding chain (IFNγ R1) deficiency was the first genetic etiology of susceptibility to mycobacteria to be identified and was initially identified in two kindreds (94, 95). Seven more families have since been reported (43, 44, 96–99b). Two kindreds with complete IFNγ receptor signaling chain (IFNγ R2) deficiency have also been reported (90, 100). The recessive IFNGR1 and IFNGR2 mutations identified are null, as they preclude cell surface expression of the receptor. A lack of cellular responses to IFNγ in vitro has been demonstrated. Four other unrelated families have recently been identified in which children with complete IFNγ R1 deficiency were found to have normal expression of IFNγ R1 molecules on the cell surface (101, 102). The mutations in IFNGR1 in these children were loss-of-function mutations because they prevented the binding of the encoded surface receptors to their natural ligand, IFNγ . In all forms of complete IFNγ R deficiency, levels of circulating IFNγ in the blood are highly elevated (103). Disseminated disease owing to BCG and/or EM, including slow-growing (M. avium, M. kansasii, and M. szulgai) and fast-growing (M. chelonae, M. abscessus, M. peregrinum, M. smegmatis, and M. fortuitum) species, were diagnosed in all patients with IFNγ R deficiency. Remarkably, M. smegmatis (44) and M. peregrinum (43) are among the least virulent mycobacteria, and infection with these bacteria had never before been reported to cause disseminated disease in humans. All BCG-vaccinated patients developed BCG disseminated disease. In all cases, EM infections occurred before the age of 3 years. No mature mycobacterial granulomas were seen. No other opportunistic infections were observed, and the course of infections owing to common childhood pathogens was unremarkable, with the exception of Salmonella enterica (94), Listeria monocytogenes (97), and a few viruses (99a, 104, 105), each reported in one patient only with the exception of cytomegalovirus in two children. Mycobacterial (BCG and EM) infections resulted in the death of about half the patients and required continuous antimycobacterial treatment in the survivors. One child recently died of tuberculosis (see More Virulent Mycobacterial Species). Complete IFNγ R deficiency thus results in susceptibility to early-onset, disseminated, and diverse BCG/EM infection, implying that IFNγ is essential for protective immunity to mycobacteria. That IFNγ R-deficient children’s vulnerability to mycobacteria is more pronounced than that of children with SCID and EDA-ID suggests that residual IFNγ -mediated immunity in the latter two conditions probably accounts for the milder course of mycobacterial disease. PARTIAL RECESSIVE IFNγ R1 AND IFNγ R2 DEFICIENCY Two siblings with partial, rather than complete, IFNγ R1 deficiency have also been reported (106). A homozygous recessive missense mutation causing an amino-acid substitution in the

2 Feb 2002

11:3

Annu. Rev. Immunol. 2002.20:581-620. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

592

AR

AR152-19.tex

CASANOVA

¥

AR152-19.SGM

LaTeX2e(2001/05/10)

P1: GJC

ABEL

extracellular domain of the receptor was identified. The receptor is expressed on the cell surface. Cells from children with partial IFNγ R1 deficiency and transfected cells carrying the mutant allele respond to IFNγ , but only at high concentrations. The missense IFNGR1 mutation probably reduces the affinity of the encoded receptor for its ligand, IFNγ . A patient with partial, as opposed to complete, IFNγ R2 deficiency was also reported (107). A homozygous nucleotide substitution was found in IFNGR2, causing a single amino-acid substitution in the extracellular region of the encoded receptor. Membrane-bound IFNγ R2 molecules were detected on the patient’s cells. The response of the patient’s cells to stimulation with IFNγ was impaired but not abolished. Transfection with the wild-type IFNGR2 gene restored full responsiveness to IFNγ . Thus, there is a causal relationship between the IFNGR2 missense mutation and weak cellular responses to IFNγ . The molecular mechanism underlying this condition remains to be determined. One IFNγ R1-deficient child had disseminated BCG and Salmonella enteritidis infections with a favorable outcome. His sibling, who had not been vaccinated with BCG, had curable symptomatic primary tuberculosis (see More Virulent Mycobacterial Species). Both are currently well at 17 and 20 years of age, with no treatment. The IFNγ R2-deficient patient with a history of BCG and M. abscessus infection is now well at 22 years of age. The clinical phenotype of the three patients with partial recessive IFNγ R deficiency is milder than that of children with complete IFNγ R deficiency. They had well-circumscribed and differentiated tuberculoid BCG granulomas. Thus, there is a correlation between the IFNGR1 and IFNGR2 genotype (loss-of-function or hypomorphic mutation), the cellular phenotype (complete or partial defect of response to IFNγ ), the histological phenotype (lepromatous or tuberculoid granulomas), and the clinical phenotype (poor or favorable outcome). The level of human IFNγ -mediated immunity seems to be the crucial factor determining the pathological lesions associated with, and the clinical outcome of, mycobacterial infections (108). PARTIAL DOMINANT IFNγ R1 DEFICIENCY Patients from 12 unrelated kindreds were found to have a dominant form of partial IFNγ R1 deficiency (86). These patients have a heterozygous small frameshift deletion in IFNGR1, downstream from the segment encoding the transmembrane domain. An interesting genetic feature of this disorder is that position 818 of IFNGR1 is the first small deletion hotspot to be identified in the human genome. Four families, each with a different dominant mutation, and five other families with a mutation at position 818, were subsequently identified (90, 105, 109, 109a,b,c). All the mutant alleles encode truncated receptors that reach the cell surface, bind IFNγ with normal affinity, dimerize and form a tetramer with two IFNγ R2 molecules, but do not transduce IFNγ -triggered signals owing to the lack of intracellular binding sites for the cytosolic molecules involved in the signaling cascade. The receptors also accumulate at the cell surface owing to the lack of an intracellular recycling site, thereby exerting a dominantnegative effect. Most IFNγ R1 dimers in heterozygous cells are nonfunctional, as they contain at least one defective molecule. The few wild-type IFNγ R1 dimers

2 Feb 2002

11:3

AR

AR152-19.tex

AR152-19.SGM

LaTeX2e(2001/05/10)

P1: GJC

Annu. Rev. Immunol. 2002.20:581-620. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

HUMAN GENETICS OF MYCOBACTERIA

593

that form in response to IFNγ account for the defect being partial rather than complete. Patients with partial dominant IFNγ deficiency are generally vulnerable to BCG; EM infections, mostly with M. avium, are frequent, generally occurring after the age of three years and occasionally in adulthood. Severe disease owing to Histoplasma capsulatum (86) and varicella zoster virus (105) were each diagnosed in one patient. One parent who died of tuberculosis was not genetically tested owing to a lack of genetic material (see More Virulent Mycobacterial Species). The prognosis of these patients is relatively good, as only three of the reported patients died, and the survivors are all well without treatment. BCG granulomas are invariably tuberculoid in these patients. The clinical features of patients with partial dominant IFNγ R1 deficiency are clearly milder than those of patients with complete deficiencies and somewhat more severe than those with partial recessive IFNγ R deficiency. This is consistent with the respective cellular phenotypes and suggests that human IFNγ -mediated immunity is a genetically controlled quantitative trait that determines the outcome of mycobacterial invasion (108). It appears that no alternative immunological pathways can partially or completely compensate for the decrease or total lack of IFNγ signaling in response to a mycobacterial challenge. PARTIAL STAT-1 DEFICIENCY Two kindreds with the same heterozygous mutation in STAT1 causing partial dominant STAT-1 deficiency have recently been described (87). STAT-1 is a critical transducer of IFN-mediated signals, either as STAT-1 homodimers, designated gamma-activating factor (GAF), or as STAT-1/STAT-2/p48 trimers, known as interferon-stimulated gamma factor 3 (ISGF3). This heterozygous STAT1 mutation decreased cellular responses to both IFNγ and IFNα in terms of the activation of GAF, but not ISGF3. The mutation results in a loss of function for both cellular phenotypes but is dominant for GAF and recessive for ISGF3 activation in the patients’ heterozygous cells stimulated with IFNs. Clinically, one patient suffered from disseminated BCG infection with tuberculoid granulomas, whereas the other had disseminated M. avium infection. They are now 36 and 10 years old and well. The clinical and cellular phenotypes of the patients were similar to those of patients with partial recessive IFN-γ R deficiency, in terms of mycobacterial disease and GAF activation. This further documents the strict genotype-phenotype correlation in the IFNγ signaling pathway, with complete lack of response to IFNγ in vitro associated with a severe clinical outcome in vivo, and partial lack of response to IFNγ associated with a good clinical outcome (108). Moreover, this observation implies that human IFNγ -mediated mycobacterial immunity is dependent on STAT-1 and GAF. The lack of severe viral illness in the patients suggests that IFN-mediated viral immunity is STAT-1–independent and/or ISGF3-dependent. COMPLETE IL-12p40 DEFICIENCY A kindred with a loss-of-function recessive mutation in the IL12B gene encoding the p40 subunit of IL-12 was reported in 1988 (110). Five other kindreds with IL-12p40 deficiency have recently been identified

2 Feb 2002

11:3

Annu. Rev. Immunol. 2002.20:581-620. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

594

AR

AR152-19.tex

CASANOVA

¥

AR152-19.SGM

LaTeX2e(2001/05/10)

P1: GJC

ABEL

(111). Neither monocytes nor dendritic cells were capable of secreting IL-12 upon stimulation. The lymphocytes of these patients secreted less IFNγ than do those of normal individuals. Impaired IFNγ secretion was successfully complemented in a dose-dependent manner by treatment with exogenous recombinant IL-12. This suggests that IFNγ deficiency is not a primary event but a consequence of inherited IL-12 deficiency. Another kindred with impaired, but not abolished, IL-12 production has also been reported (83, 85). The genetic defect was not identified, but the familial pedigree suggests recessive X-linked inheritance rather than autosomal recessive IL-12p40 deficiency. All patients but one with complete IL-12 deficiency owing to IL12B mutations had BCG infection, associated in only one child with EM infection (M. chelonae) and in another with tuberculosis (see More Virulent Mycobacterial Species); one third had S. enterica infections (which was the only infection in one child), and one had Nocardia asteroides infection. Only five children died of infection, and the survivors are all well without treatment. Thus, IL-12–deficient children probably suffer from mycobacterial infection primarily because their IFNγ -mediated immunity is impaired. Residual, IL-12–independent secretion of IFNγ probably accounts for the clinical phenotype being milder than that of children with complete IFNγ R deficiency. The clinical outcome of IL-12– deficient patients, however, varies from case to case. Mutations in the IL12RB1 gene encoding the β1 subunit of the IL-12 receptor were initially identified in six kindreds (112, 113). Three additional families were subsequently identified (114–116). All patients were homozygous for recessive mutations precluding the surface expression of IL-12Rβ1, and IFNγ secretion in vitro by otherwise functional NK cells and T cells was impaired (112, 113, 116a). The molecular complementation of defective cells by transfection with the wild-type IL12RB1 gene has recently been reported, confirming the pathogenic role of missense mutations in two kindreds (115, 116). A heterozygous IL12RB1 missense mutation was identified that may have contributed to another patient’s predisposition to mycobacteriosis (116b). A patient with a related phenotype of vulnerability to mycobacterial and staphylococcal disease was found to respond poorly to IL-12 despite having normal IL-12Rβ1 (117). The clinical phenotype of IL-12Rβ1–deficient patients appears to be similar to that of IL12p40–deficient children, suggesting that IL-12Rβ1–independent IL-12 signaling (118) has little impact on immunity to mycobacteria. BCG infections were curable, and EM infections (M. chelonae in one child, M. avium in the other cases) occurred only after the age of three years, with three cases diagnosed in adulthood. Half the patients had associated S. enterica infections, but no other infections were reported. One patient presented with abdominal tuberculosis (see More Virulent Mycobacterial Species). Only one patient died of EM infection, and the other patients were well at the last follow-up. The histological phenotype also appears to be milder, as BCG granulomas were found to be well delimited and well differentiated. Unlike the case of defects of cellular responses to IFNγ , there seems to be significant interfamilial and intrafamilial clinical heterogeneity, as two patients with IL-12Rβ1 DEFICIENCY

2 Feb 2002

11:3

AR

AR152-19.tex

AR152-19.SGM

LaTeX2e(2001/05/10)

P1: GJC

Annu. Rev. Immunol. 2002.20:581-620. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

HUMAN GENETICS OF MYCOBACTERIA

595

IL-12Rβ1 deficiency were resistant to BCG (114, 115), one of whom did not even develop atypical mycobacteriosis (115). She was vaccinated three times with live BCG and did not show any adverse effect, whereas her brother had disseminated BCG-osis (115). Accordingly, EM disease generally presented in childhood but occurred in adulthood in three patients (113, 116) and we recently identified asymptomatic affected siblings in several kindreds (unpublished data). The occurrence of strictly asymptomatic individuals with genetic mutations abolishing IL-12– mediated immunity is a conceptually challenging observation that should perhaps lead to the revision of certain immunological dogmas (119). Unlike with IFNγ R deficiency there is apparently no correlation between the IL12B and IL12RB1 genotypes and clinical phenotype. There may be alternative pathways that compensate for the loss of IL-12 signaling and result in different cellular phenotypes (e.g., different levels of production of IFNγ ) in the patients. This suggests that IL-12 is one of several inducer cytokines and IFNγ is the only effector cytokine involved in immunity to mycobacteria.

Immunological Issues and Genetic Implications The identification of Mendelian defects predisposing patients to BCG/EM disseminated disease has revealed that T cells, NEMO-dependent NF-κB activation, and IL-12–dependent, STAT-1–mediated activation by IFNγ are crucial for protective immunity to mycobacteria. The clinical course of mycobacterial infection in patients with anhidrotic ectodermal dysplasia with immunodeficiency, lymphedema, and osteopetrosis (OL-EDA-ID) and complete IFNγ R deficiency is more severe than that in severe combined immunodeficiency (SCID) patients, which is itself more severe than that in patients with other forms of EDA-ID and Mendelian susceptibility to mycobacterial disease (MSMD). Among patients with MSMD, the level of IFNγ -mediated immunity determines the severity of mycobacterial disease. This suggests that impaired IFNγ -mediated immunity is the principal pathogenic mechanism underlying mycobacterial disease in patients with MSMD, EDA-ID, and SCID and that quantitative variations in production of, or response to, IFNγ between the different types of defect account for the differences in clinical phenotype. The identification of novel molecular defects in patients with PIDs and mycobacterial disease should enable a genetic dissection of the immunological pathways connected with IFNγ -mediated immunity. The identification of novel molecular defects upstream of IL-12 and downstream of STAT-1 in patients with MSMD should facilitate a genetic dissection of the principal axis of human immunity to mycobacteria. A lack of development of T cells results in a broad vulnerability to viruses, bacteria, and fungi, and impaired NF-κB activation results in a vulnerability to different types of bacteria. In contrast, patients with an inherited defect of the IL-12-IFNγ axis show a vulnerability that is mostly restricted to mycobacteria (120, 121). Infections by S. enterica in less than half the patients, and by H. capsulatum, L. monocytogenes, N. asteroides, and a few viruses in only one patient

2 Feb 2002

11:3

Annu. Rev. Immunol. 2002.20:581-620. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

596

AR

AR152-19.tex

CASANOVA

¥

AR152-19.SGM

LaTeX2e(2001/05/10)

P1: GJC

ABEL

each, somewhat qualify this conclusion and suggest that other intracellular microorganisms may occasionally threaten these patients. The observation that IL12(R)–, IFNγ R–, and STAT-1–deficient patients are vulnerable to mycobacteria was expected, based on the mouse model (35, 120), but their apparent resistance to most other micro-organisms is surprising. In natural conditions of infection, it appears that the IL-12-IFNγ axis exerts some form of specificity for protective immunity to mycobacteria. This observation may reflect in part an ascertainment bias, and the molecular and clinical investigation of patients with various infectious diseases, living in different environments, is required. The lack of T cells obviously accounts for mycobacterial disease in SCID patients. In patients with EDA-ID and MSMD the cells responsible for vulnerability to mycobacteria are unknown. As IFNγ is a major macrophage-activating cytokine, macrophages probably play a key role in the pathogenesis of mycobacterial infections in patients with IFNγ R and STAT-1 deficiency. However, lymphocytes and other immune cell subsets may also be involved, either directly, as they express IFNγ R molecules, or indirectly, because impaired macrophage activation by IFNγ may restrict their activation by monokines (80, 98). Along these lines, the cells responsible for mycobacterial disease in IL-12p40 deficiency may be dendritic cells and/or macrophages, and those responsible in IL-12Rβ1 deficiency may be NK and/or T cells. Finally, NEMO and NF-κB are ubiquitous and triggered by different stimuli, raising questions about which pathways and which cells are responsible for the phenotype. Antigen-presenting cells, expressing the Toll-like receptors and the TNFα receptor, along with lymphocytes, expressing the IL-1β and IL-18 receptors, are good candidates. The development of conditional knock-out mice will be necessary to address the question of the cellular basis of mycobacterial disease in patients with impaired NF-κB or IFNγ immunity.

MORE VIRULENT MYCOBACTERIAL SPECIES (M. TUBERCULOSIS, M. LEPRAE, M. ULCERANS) Epidemiological Studies Factors known to contribute to the development of clinical tuberculosis include nongenetic host factors, such as acquired immunodeficiency (e.g., AIDS) or acquired immunity (e.g., BCG vaccination), and environmental factors, such as microbial virulence (e.g., resistance to antibiotics) and social conditions (e.g., poverty) (Figure 1) (14, 15, 122). The same factors, with the apparent exception of HIV infection and immunodeficiency, affect the development of leprosy (19, 20). Less is known about the factors that influence the development of Buruli ulcers (17). Many epidemiological studies have indicated that the host’s genetic makeup plays a role in the considerable variability of clinical response to infection with M. tuberculosis and M. leprae (123). Several studies have shown that an individual’s level of resistance to M. tuberculosis infection is correlated with the geographical origin of the family, with the most vulnerable individuals tending to have

2 Feb 2002

11:3

AR

AR152-19.tex

AR152-19.SGM

LaTeX2e(2001/05/10)

P1: GJC

Annu. Rev. Immunol. 2002.20:581-620. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

HUMAN GENETICS OF MYCOBACTERIA

597

ancestors originating from areas once free of tuberculosis (124). The incidence of tuberculosis is particularly high during outbreaks in populations with no ancestral experience of the infection, such as native Americans (125). Black populations have also been shown to be more susceptible than Caucasians to tuberculosis (124, 126). Familial aggregation studies have also provided convincing evidence. Twin studies have shown much higher concordance rates for monozygotic than dizygotic pairs for both clinical tuberculosis (127) and leprosy (128). Finally, several segregation analyses (129) have clearly shown that susceptibility to leprosy has a significant genetic component. In particular, a segregation analysis performed on Desirade Island in the French West Indies found evidence for a recessive major gene controlling susceptibility to leprosy per se (i.e., regardless of clinical subtype) (130). To our knowledge, the role of host genetic factors in M. ulcerans infection has yet to be investigated (17). The following sections therefore deal with the principal genetic studies carried out for tuberculosis and leprosy.

Mendelian Disorders PRIMARY IMMUNODEFICIENCIES Virtually nothing is known regarding the relationships between any Mendelian disorder and infection with M. leprae or M. ulcerans. In contrast, a number of children with Mendelian disorders, including cystic fibrosis and primary immunodeficiencies (PIDs), have been found to be highly vulnerable to M. tuberculosis and related species (46). This was shown for one child with SCID (131), two patients with X-linked hyper IgM syndrome associated with CD154 mutations (132), one of 368 CGD patients from North America (133) and six of seven CGD patients from Hong Kong (134), where tuberculosis is endemic, and two EDA-ID patients including one child with a NEMO mutation (76, 135). The underlying immunodeficiency probably favored the development of tuberculosis in these patients because these types of immunodeficiency are known to predispose patients to infection with less virulent mycobacteria (except CD154 deficiency; see Mildly Virulent Mycobacteria), and the course of tuberculosis was invariably severe and often fatal. Two adults with tuberculosis were reported to suffer from common variable immunodeficiency (136) and mild T-cell deficiency (137). However, a causal relationship cannot be demonstrated in these two cases, as tuberculosis may have been coincidental. The susceptibility to mycobacterial infections observed in these conditions demonstrates that T cells, T cell–associated CD154, phagocytic NADPH oxidase, and NEMO-dependent NF-κB activation are essential components of immunity to M. tuberculosis. It is not entirely clear, however, which PIDs predispose patients to infection with virulent mycobacteria because exposure and infection are generally not detected and are relatively rare in countries in which PID diagnosis is possible, whereas constant exposure and infection are likely to occur in endemic countries where these PIDs are largely underdiagnosed. In Hong Kong, an endemic region where PIDs are well diagnosed, most patients with CGD suffer from severe tuberculosis (134). Thus, the number and type of PIDs predisposing patients to severe disease are currently underestimated, and more studies in endemic areas are

2 Feb 2002

11:3

598

AR

AR152-19.tex

CASANOVA

¥

AR152-19.SGM

LaTeX2e(2001/05/10)

P1: GJC

ABEL

Annu. Rev. Immunol. 2002.20:581-620. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

required to document more cases. It is therefore less clear which genes and immune pathways are essential for protective immunity against M. tuberculosis than BCG and EM. MENDELIAN DISORDERS OF THE IL-12-IFNγ AXIS Patients with MSMD are vulnerable to M. tuberculosis. The classical phenotype of these patients is vulnerability to poorly virulent mycobacterial species such as Bacille Calmette-Gu´erin (BCG) and environmental mycobacteria (EM), but five patients developed tuberculosis, which in three cases was the sole clinical manifestation of their underlying genetic disorder. One child with IL-12p40 deficiency (111) and another with complete IFNγ R1 deficiency (unpublished data) had curable BCG-osis and died of tuberculosis. The reason only two patients with an inherited defect of the IL-12-IFNγ axis and BCG/EM clinical disease also had tuberculosis is probably that most patients were not exposed to M. tuberculosis. One parent of a child with partial dominant IFNγ R1 deficiency died of tuberculosis, but no material was available for genetic analysis (86). One child who was not vaccinated with BCG developed a symptomatic form of primary tuberculosis (106). The child was diagnosed with partial recessive IFNγ R1 deficiency because her older brother had symptomatic BCG-osis. Otherwise, she would have probably remained one in a multitude of cases of symptomatic primary tuberculosis, as she suffered no other unusual infections, including other mycobacterial infections. Finally, another patient who was vaccinated three times with BCG with no adverse effect (and can be considered as being truly resistant to BCG, as the tuberculin skin tests were positive, implying that live BCG had indeed been inoculated) and who did not develop atypical mycobacteriosis presented with full-blown abdominal tuberculosis at 18 years of age (115). This observation suggests that susceptibility to severe forms of tuberculosis, such as extra-pulmonary tuberculosis, may be caused in BCG-resistant and otherwise healthy individuals by purely Mendelian disorders of the IL-12-IFNγ axis (138). The precise frequency of such types of Mendelian predisposition to tuberculosis is unknown.

Complex Inheritance In more common situations, the strategy used to investigate the genetic component of the response to M. tuberculosis and M. leprae is based on population genetic epidemiology studies (139). The ultimate goal of these analyses is to identify the genes, and the alleles of these genes, that significantly account for the phenotype of interest (e.g., affected/unaffected) and possible interactions of these alleles with environmental risk factors. Recent developments, such as the establishment of a genetic map of the human genome based on highly polymorphic markers (140), the growing availability of intragenic single nucleotide polymorphisms (141, 142), and the sequencing of the human genome (22, 23) have led to the creation of tools essential for these genetic studies. Numerous methods have been (and are being) developed. They generally fall into two categories (143, 144): linkage analysis methods, which seek to locate a chromosomal region that segregates nonrandomly

2 Feb 2002

11:3

AR

AR152-19.tex

AR152-19.SGM

LaTeX2e(2001/05/10)

P1: GJC

Annu. Rev. Immunol. 2002.20:581-620. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

HUMAN GENETICS OF MYCOBACTERIA

599

with the phenotype of interest within families, and association studies, which test for a significant association between a specific genetic polymorphism and a phenotype within a population (Figure 2). One major advantage of linkage studies is that they can be used to explore the whole genome (genome screen) ensuring that all major loci involved in the control of a phenotype are identified and making it possible to discover new genes (and, potentially, new immunological pathways). Association studies investigate the role of polymorphisms (or alleles) of candidate genes defined on the basis of either their location (in regions identified by linkage analysis) or function (genes involved in the immune response to mycobacteria). After presenting the main candidate genes that have been investigated by linkage and/or association studies in both diseases, we review the two genome screens recently reported in tuberculosis and leprosy (Table 2). THE MAJOR HISTOCOMPATIBILITY COMPLEX Human leukocyte antigen (HLA) molecules are highly polymorphic and present antigenic peptides to α/β T cells. This has led to extensive studies of the role of the HLA genes in tuberculosis and leprosy. Most of the studies carried out were population-based association surveys in adults, comparing HLA class I and/or class II alleles in unrelated cases and unrelated controls. No consistent findings were reported for HLA-I alleles. Several studies have reported a higher frequency of HLA-DR2 in patients with pulmonary tuberculosis (PTB) from Indonesia (145) and India (146, 147), with an associated odds ratio (OR) of 1.8–2.7. If the prevalence of the disease is lower than 10%, which is generally the case for tuberculosis, the OR is a valid estimate of the relative risk (i.e., the risk of disease with a particular genotype versus the risk of disease without this genotype). A family-based association study also provided evidence that HLA-DR2 is involved in PTB (148), by showing a skewed transmission of DR2 (around 80%) to affected offspring from DR2 heterozygous parents in a sample of 25 multiple-case Indian families. Other case-control studies failed to replicate the HLA-DR2 association in Chinese (149), Mexican (150), and Indian (151) populations. Nevertheless, the use of HLA serologic techniques has not provided an accurate resolution of HLA class II types (152). Recent studies with molecular DNA-based typing methods have reported a high frequency of DRB1∗ 1501 (a DR2 allele) in PTB patients from India (153, 154) and Mexico (155), with estimated ORs of 2.7–8. Two DQ1 alleles, DQB1∗ 0503 in Cambodia (156) and DQB1∗ 0501 in Mexico (155), were also found to be associated with PTB. HLA has been reported to be involved in tuberculoid and lepromatous leprosy, but not in leprosy per se (i.e., all clinical forms of leprosy). Case-control association studies (157) have shown that HLA-DR3 frequencies are high in patients with tuberculoid leprosy and low in patients with lepromatous leprosy. A high frequency of HLA-DR2 was also reported in both tuberculoid and lepromatous patients (158). Two family-based association studies in India (159) and Egypt (160) found a skewed distribution of the DR2 allele in siblings with tuberculoid leprosy. Using molecular HLA typing, an association was found between tuberculoid leprosy and DR2 alleles DRB1∗ 1501 and DRB1∗ 1502 in India (161, 162). Sib-pair

Genome-wide linkage analysis

10p13

224 multicase families

1 large pedigree of Aboriginal Canadians 20 leprosy families

44 families

(176) (178) (179) (177) (180) (184) (199)

4.1 (1.9–9.1) 1.8 (1.1–3) 0.5 (0.3–0.7) <0.04 <2 × 10−5 <0.002 <2 × 10−5

c

Only association studies with molecular HLA typing are presented, but numerous other serological studies have shown that DR2 (DRB1∗ 1501 and DRB1∗ 1502 are DR2 alleles) is associated with pulmonary tuberculosis, tuberculoid leprosy, and lepromatous leprosy (see text).

Odds ratios with 95% confidence intervals are provided for case/control association studies, whereas p-values are shown for other studies.

b

Candidate genes and chromosomal regions are shown for association studies and linkage studies, respectively.

South India

Vietnam

Granulomatous reaction to intradermal lepromin Paucibacillary leprosy

Canada

Pulmonary tuberculosis

Guinea-Conakry

192 cases/192 controls 267 cases/202 controls

Korea Japan

INT4 C + 30 UTR del 30 UTR del 50 (GT), 203bp (homozygous) INT4 C

(163) (164) (159) (165) (160) (160) (166)

<0.005 <0.05 <0.05 <0.05 <0.0005 <0.006 <0.05

LaTeX2e(2001/05/10)

a

Linkage analysis in one candidate region

Pulmonary tuberculosis

Family-based association study

410 cases/417 controls

The Gambia

(162)

4.7 (2–11)

AR152-19.SGM

2q35 (NRAMP1 region)

Pulmonary tuberculosis

Case/control association study

NRAMP1

16 multicase families 72 multicase families 13 multicase families 28 multicase families 15 multicase families 28 multicase families 26 multicase families

(153) (155) (153) (161)

Ref.

4.8 (1–23.3) 7.9 (2.7–23.1) 2.7 (1.1–6.7) 5.7 (2–16)

OR (95% CI)/b p-value

AR152-19.tex

Lepromatous leprosy

Surinam South India Central India Venezuela Egypt Venezuela China

54 cases/44 controls

North India

DRBI 1501 DRBI∗ 1501 DRBI∗ 1502 DRBI∗ 1501 +DRBI∗ 1502 DRBI∗ 1501 +DRBI∗ 1502

∗

Alleles

¥

Tuberculoid leprosy

20 cases/46 controls 50 cases/95 controls 39 cases/46 controls 28 cases/47 controls

North India Mexico North India North India

Sample

CASANOVA

Linkage analysis in one candidate region

Tuberculoid leprosy

Pulmonary tuberculosis

Population

AR

6p21 (MHC region)

Case/control association study

Phenotype

600

HLA-DRBI

Design

11:3

c

Candidate genes/a Chromosomal region

TABLE 2 Complex predisposition to mycobacterial disease: a selection of association and linkage studies

Annu. Rev. Immunol. 2002.20:581-620. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

2 Feb 2002 P1: GJC

ABEL

2 Feb 2002

11:3

AR

AR152-19.tex

AR152-19.SGM

LaTeX2e(2001/05/10)

P1: GJC

Annu. Rev. Immunol. 2002.20:581-620. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

HUMAN GENETICS OF MYCOBACTERIA

601

linkage studies provided further evidence for the involvement of HLA in polar leprosy. Nonrandom segregation of parental HLA haplotypes was observed among sets of children with tuberculoid leprosy from Surinam (163), Central (159) and South (164) India, Venezuela (165), and Egypt (160). Similar results were obtained for siblings with lepromatous leprosy from Venezuela (165) and China (166). In contrast, HLA haplotypes segregated randomly among affected (164) and healthy (165) siblings in multicase leprosy families, suggesting that HLA has little effect on susceptibility to leprosy per se (157). This is consistent with HLA playing a role at relatively late stages of infection, with antigen-specific T-cell immunity governing the polarization of leprosy to tuberculoid or lepromatous forms. The role of other genes located within the MHC was also assessed in casecontrol studies. No association was found between PTB and a tumor necrosis factor (TNF)-α promoter polymorphism (referred to as TNF2) in populations from Cambodia (156) and Brazil (167). A high frequency of TNF2 was reported in Indian lepromatous leprosy patients with an OR of ∼2 (168, 169). In a population from North India, five polymorphisms in the transporter associated with antigenprocessing (TAP) 2 gene were identified. With respect to controls, a higher frequency of TAP2-A/F (OR = 4.3, p < 0.002) and TAP2-B (OR = 3.5, p < 0.006) was found in PTB patients and tuberculoid leprosy patients, respectively (170). Overall, the most convincing associations within the MHC have been found between susceptibility to tuberculosis and polar leprosy and HLA class II alleles. Although little is known about the molecular nature of HLA-restricted mycobacterial antigens, it has been suggested that certain HLA class II genes, by selecting certain sets of antigenic peptides and specific helper T cells, may contribute to the development of tuberculosis (156) or polar leprosy (162). THE NRAMP1 GENE Natural resistance associated macrophage protein-1 (Nramp1) was the first mycobacterial susceptibility gene (Bcg) identified in the mouse (36, 171, 38, 39). A natural mutation in Nramp1 impairs early immunity to several mycobacteria, including M. lepraemurium (172), the rodent-tropic equivalent of M. leprae. This mutation does not appear to favor M. tuberculosis infection (173), but its effects on infection with the rodent-tropic equivalent M. microti have not been tested. Nramp1 encodes an integral membrane protein that depletes phagosomes of divalent metal cations essential for microbial survival (174). NRAMP1 is therefore a candidate gene, as it is the human ortholog of Nramp1 (175). In a case-control study performed in The Gambia (176), two NRAMP1 polymorphisms were found to be independently associated with PTB. The first was in intron 4 (INT4) and the other was in the 30 untranslated region (30 UTR). Heterozygosity for both polymorphisms was associated with the highest risk of tuberculosis (OR = 4.1, p < 0.001). A family-based association study in Guinea-Conakry found a skewed transmission of the INT4, but not the 30 UTR, polymorphism to the offspring of patients with PTB (177). PTB was found to be associated with the 30 UTR polymorphism in a case-control study in Korea (178) but not in Japan (179). It is not yet clear whether these modest associations reflect the effect of the polymorphisms studied or that

2 Feb 2002

11:3

Annu. Rev. Immunol. 2002.20:581-620. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

602

AR

AR152-19.tex

CASANOVA

¥

AR152-19.SGM

LaTeX2e(2001/05/10)

P1: GJC

ABEL

of other polymorphisms in linkage disequilibrium. A linkage study performed in a large Aboriginal Canadian pedigree found a major locus of susceptibility to clinical tuberculosis in chromosomal region 2q35 (180). The most significant results (p < 2 × 10−5) were obtained with a dominant susceptibility allele, the carriers of which had a ten times higher than normal risk of contracting tuberculosis. Whether this linkage reflects the role of NRAMP1 itself or that of a closely linked gene remains to be established. In either case, this important study indicates that, at least in certain extended families only recently exposed to M. tuberculosis, subentities with Mendelian inheritance may be involved in the genetic control of tuberculosis (138). The role of NRAMP1 in leprosy has mostly been investigated through linkage studies. A sib-pair study in Vietnam showed significant linkage (p < 0.005–0.02) between leprosy per se and NRAMP1 haplotypes defined by six intragenic and four juxtagenic polymorphisms (181). Unlike the HLA region, NRAMP1 was not linked with either type of polar leprosy (lepromatous or tuberculous). This suggests that NRAMP-1 may be involved in early protective immunity against M. leprae, consistent with its known function in mice both in vitro and in vivo (38, 39). This linkage study, along with the segregation analysis performed in the same population (129), also suggested genetic heterogeneity according to the ethnic origin of the families (Vietnamese or Chinese). Such heterogeneity may account, at least in part, for the failure of two previous studies to detect a linkage between leprosy and the NRAMP1 region in families from Pakistan and Brazil (182) and French Polynesia (183). In the Vietnamese study the NRAMP1 region was also found to be linked (p < 0.002) with the in vivo Mitsuda reaction, which measures the delayed immune response against intradermally injected lepromin (184). This may reflect the previously identified association of NRAMP1 with leprosy per se, with NRAMP1 primarily involved in host defense against M. leprae. Alternatively, it may suggest that NRAMP1 is also involved in the control of delayed immune responses to lepromin. Further studies are required to discriminate between these two hypotheses. OTHER CANDIDATE GENES These studies were recently reviewed (123, 185) and are only briefly presented here. The active form of vitamin D, 1α,25-(OH)2D3, modulates the differentiation, growth, and function of various cell types, including dendritic cells (186). The influence of an exon polymorphism (T and t alleles) in the vitamin D receptor gene (VDR) has been investigated in several studies. In The Gambia a lower proportion of tt homozygotes was found among patients with PTB than among controls (OR = 0.5, p < 0.02) (187). In another case-control study, of Gujarati Asians living in England, no convincing association was found between PTB and VDR polymorphisms (188, 189). Finally, a study was performed in leprosy patients from India (190) in which tt homozygotes were more frequent in tuberculoid leprosy patients than controls (OR = 3.2, p < 0.001) and TT homozygotes were more frequent among lepromatous leprosy patients than among controls (OR = 1.7, p < 0.04).

2 Feb 2002

11:3

AR

AR152-19.tex

AR152-19.SGM

LaTeX2e(2001/05/10)

P1: GJC

Annu. Rev. Immunol. 2002.20:581-620. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

HUMAN GENETICS OF MYCOBACTERIA

603

Mannose-binding lectin (MBL) is a serum protein produced by the liver that binds microbial surfaces and activates the complement cascade (191). The role of loss-of-function MBL polymorphisms in tuberculosis has been investigated in three case-control studies. In India the frequency of individuals homozygous for MBL mutations was higher in PTB patients than controls (OR = 6.5, p < 0.009) (192). However, although no association was found in The Gambia (193), one of the MBL mutant alleles was found to be protective against PTB in South Africa (OR = 0.4, p < 0.02) (194). The results of the Indian study suggest that MBL plays a beneficial role in host immunity (191), whereas the South African study suggests that low MBL levels may protect against certain infectious agents (195). The proinflammatory cytokine IL-1β plays an important role in mouse defenses against mycobacteria, as suggested by studies in IL-1 receptor-deficient mice (196). The role of polymorphisms in the genes encoding IL-1β and IL-1 receptor antagonist (IL-1Ra) was recently investigated in a population of Gujarati Asians living in England (197). Each healthy subject carrying the IL1RA allele A2 produced 1.9 times the IL-1Ra produced by the remaining subjects. No differences in the frequency of IL1B and IL1RA polymorphisms were observed between patients with tuberculosis and controls. However, subjects carrying a certain IL1B allele and not the IL1RA A2 allele were overrepresented among patients with tuberculous pleurisy (p < 0.03). Finally, the IL1RA A2 allele was found to be associated with a reduced Mantoux response to purified protein derived from M. tuberculosis. Altogether, these studies do not point to a major role of mutations in VDR, MBL, IL1B, and IL1RA in human tuberculosis or leprosy, and further studies are awaited. GENOME-WIDE SCREENS The two genome-wide screens performed for tuberculosis and leprosy were based on affected sib-pair linkage studies, which test whether affected siblings share more parental alleles in some chromosomal regions than would be expected assuming the random transmission of alleles from parents to child. In genome-wide screens convincing evidence for linkage is defined as a lod-score >+3 (i.e., a p-value <10−4), owing to the problem of multiple testing with numerous markers. This immunologically blind strategy should have the advantage of identifying new genes. The first study concerned adult PTB in The Gambia (85 sib-pairs) and South Africa (88 sib-pairs) (198). No region of the genome showed convincing evidence for linkage, ruling out the presence of a major susceptibility gene in these families. Two regions located on chromosomes 15q (p < 0.001) and Xq (p < 0.005) provided limited evidence suggestive of linkage. Interestingly, no linkage was found with the regions containing the NRAMP1 and VDR genes, for which associations were previously reported in the Gambian population (176, 187), indicating that these genes had no major effect on PTB in this population. The second genome-wide screen concerned tuberculoid/paucibacillary leprosy in a total of 245 affected sib-pairs from South India (199). Significant evidence was obtained for a linkage (p < 2 × 10−5) with chromosomal region 10p13,

2 Feb 2002

11:3

604

AR

AR152-19.tex

CASANOVA

¥

AR152-19.SGM

LaTeX2e(2001/05/10)

P1: GJC

ABEL

indicating the existence of a major susceptibility gene for paucibacillary leprosy in this population. Surprisingly, no linkage was observed with the HLA region, despite evidence provided by sib-pair studies in several countries, including South India (164), for linkage with paucibacillary leprosy.

Annu. Rev. Immunol. 2002.20:581-620. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

Genetic Issues and Immunological Implications It is important to recall the inherent limitations of population-based association studies (185). A major methodological problem is multiple testing (e.g., in every 20 tests 1 will be significant at the 0.05 level owing to chance alone). A key methodological point that is often overlooked is to correct the observed p-values for the number of tests performed (and not only the number published). A second problem relates to linkage disequilibrium. A polymorphism associated with a disease may not necessarily be disease-causing but may be in linkage disequilibrium with a genuine disease-causing polymorphism. Conversely, the absence of association with a polymorphism of a given gene does not exclude the possibility that that gene is involved, as other polymorphisms may also have an effect. Finally, a third critical point in case-control studies is the choice of controls. They must be unrelated (like the cases) and should be as similar as possible to the cases in terms of ethnic origin, age, and ideally, in terms of exposure to the infectious agent. The misspecification of some of these points or the presence of a population admixture can markedly affect both the robustness and the power of the analysis. Familybased association studies overcome control selection bias. Linkage studies share with association studies the issue of multiple testing but avoid the other problems. However, they are generally less powerful than association studies (when the allele tested is disease-causing) and can only detect major genes. Although linkage studies can demonstrate the involvement of a chromosomal region, they are not useful for very fine mapping of the genes responsible for complex phenotypes such as tuberculosis or leprosy. The principal convincing results obtained in previous studies, taking into account these potential limitations, are summarized in Table 2. They are (a) the association of HLA-DR2 (or certain molecular subtypes) with PTB and polar forms of leprosy; (b) the role of NRAMP1 (or a closely linked gene), which may be a major susceptibility gene to tuberculosis in certain populations; and (c) the presence of a major locus on chromosome 10p predisposing patients to paucibacillary leprosy. The role of HLA-DR2 prompts one to search for an immunological basis for this genetic observation. The identification of the latter two disease-causing genes may reveal novel mechanisms of immunity to mycobacteria. Most other findings require confirmation in additional genetic studies before immunological investigations are undertaken to validate fully the role of candidate genes and alleles. Genome-wide screens have been undertaken by various research groups in other populations and may corroborate the previous results or lead to the identification of novel major genes. It is also clear that only a minority of candidate genes have yet been tested and many others are potentially relevant, such as those involved

2 Feb 2002

11:3

AR

AR152-19.tex

AR152-19.SGM

LaTeX2e(2001/05/10)

P1: GJC

HUMAN GENETICS OF MYCOBACTERIA

605

Annu. Rev. Immunol. 2002.20:581-620. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

in Mendelian predisposition to mycobacterial disease (see Mildly Virulent Mycobacteria) and/or in the numerous pathways identified in the mouse that influence the immune response to mycobacteria (35). In any case, for the effect of a given allele to be convincing, especially when it is thought to have a moderate effect, it must be validated by immunological studies. This will be difficult because the relative risks of disease are low, implying that the biological effect of each pathogenic allele is mild.

CONCLUSION The investigation of rare Mendelian predisposition to mycobacteria has demonstrated that T cells, NF-κB signaling, and the IL-12-IFNγ axis play a crucial role in human immunity to mycobacteria (Table 1). Other molecules involved, such as T-cell CD154 and phagocyte NADPH oxidase, have been shown to be less important. Natural mutations are likely to occur in almost all human genes; therefore most, if not all, Mendelian defects predisposing to severe mycobacterial disease will be identified, provided that most patients are thoroughly investigated. In addition to candidate genes in known pathways, positional mapping may identify new genes and pathways involved in immunity to mycobacteria. Studies of complex predisposition in the general population have led to fewer immunological advances but hold promise of interesting discoveries (Table 2). They suggest a role for NRAMP1 and HLADRB1 in leprosy and tuberculosis, although further evidence is required for definitive conclusions to be drawn. Other candidate genes should be tested, including those controlling the IL-12-IFNγ axis. Complex predisposition also involves major genes on chromosomes 2q and 10p. Novel segregation studies and genome-wide screens should be carried out in the future. The elucidation of major mycobacterial susceptibility genes, primarily through positional cloning approaches, will be of immunological interest. An important question arises as to whether there are “public” genes that control immunity to the diverse mycobacterial species (weakly virulent BCG/EM, more virulent M. leprae and M. tuberculosis) (138). A single gene may be involved in the three levels of genetic control, with rare mutations being responsible for Mendelian susceptibility to BCG/EM (and more virulent species upon exposure), relatively rare variants having a major gene effect and more common polymorphisms having a milder effect on the risk of clinical tuberculosis or leprosy. The correlation observed between the degree to which IFNγ -mediated immunity is impaired and the severity of BCG/EM disease is consistent with this hypothesis. The demonstration that IFNGR1, IL12B, and IL12RB1 cause Mendelian predisposition not only to BCG and EM but also to M. tuberculosis is also consistent with this hypothesis. However, mycobacteria display such a high level of diversity, both between and within species, in terms of habitat, metabolic requirements, speed of growth, mode of infection, cell and tissue tropism, survival within phagocytes, and pathogenicity, and there are probably “private” susceptibility genes, the identification of which

2 Feb 2002

11:3

Annu. Rev. Immunol. 2002.20:581-620. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

606

AR

AR152-19.tex

CASANOVA

¥

AR152-19.SGM

LaTeX2e(2001/05/10)

P1: GJC

ABEL

should cast light on specific aspects of immunity to mycobacteria. Dissecting the genetic control of public and private immunity to mycobacterial species will be an important aspect of future research. The question also arises as to whether it is possible for human beings, as individuals, kindreds, or populations, to have the same phenotype of vulnerability to a given mycobacterial species and yet have different susceptibility genotypes. In other words, is there only one or are there several forms of genetic susceptibility to each Mycobacterium species? Mendelian vulnerability to weakly virulent species is known to be genetically heterogeneous, with mutations in IFNGR1, IFNGR2, STAT1, IL12B, and IL12RB1 being responsible for susceptibility to BCG and EM disease (not to mention the PIDs, which predispose to mycobacterial and other infections). Thus, vulnerability to the more common mycobacteria is also likely to show nonallelic genetic heterogeneity, both in its Mendelian and complex components. The different histories of discrete populations and geographical and ecological diversity also suggest that the genes controlling human immunity have evolved differently in different regions of the world (200). It will be interesting to determine why particular individuals or populations are vulnerable to one particular mycobacterial species. There may be susceptibility genes, mutations of which cause vulnerability in all populations/individuals studied, and other susceptibility genes that have an impact in certain individuals/populations only. These studies may have important immunological implications. The human model is also ideal for dissection of the various steps from exposure to clinical disease (Figure 1). The genes that control the different clinical and immunological phenotypes may differ from each other, even though some genes are expected to be involved in all steps of immunity to mycobacteria. Most studies have focused on simple clinical phenotypes, such as affected versus nonaffected, implying that protective immunity was under study. The investigation of more subtle clinical phenotypes, such as pulmonary and extra-pulmonary tuberculosis, may reveal important aspects of immunity. It will be difficult, if not impossible, to investigate the genetic control of innate immunity to mycobacteria, as it is difficult both to assess exposure and to exclude strictly any involvement of adaptive immunity. Nevertheless, studies dealing with immunological phenotypes (detectable after infection has stimulated adaptive immunity), such as T-cell mediated hypersensitivity reaction or specific antibody titer, may reveal the genetic control of T-cell and B-cell responses to mycobacteria. There is also a need to investigate clinical and immunological phenotypes in different populations and in response to different mycobacterial species. A comprehensive genetic dissection of the different steps of immunity to mycobacteria in natural conditions of infection, hence in natural conditions of immunity, is possible in the human model. The past 50 years have seen the development and success of the mouse model of mycobacterial infection (35). The inherent limitations of this model should lead to consideration of the human model as a complementary approach. Human genetics of mycobacterial diseases largely relies on the principles of immunity to mycobacteria established in the mouse model and will in turn stimulate

2 Feb 2002

11:3

AR

AR152-19.tex

AR152-19.SGM

LaTeX2e(2001/05/10)

P1: GJC

Annu. Rev. Immunol. 2002.20:581-620. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

HUMAN GENETICS OF MYCOBACTERIA

607

important studies in the mouse. More generally, the human model can be applied to all microbial genera or species, provided human vulnerability to the organism of interest is the exception rather than the rule. For highly virulent microorganisms, the human model is useful for unraveling microbial pathogenesis rather than host immunity, as it facilitates the genetic dissection of resistance to the organism rather than vulnerability. Regardless of its obvious and important medical implications and from a strictly immunological standpoint, the human model offers the unique advantage of investigating the function of immune genes not only in vivo, but also in natural conditions of infection. The human model may thus become a model of choice for future research in immunology. Understanding the molecular basis of protective immunity to microorganisms in natural conditions of exposure is certainly one of the most fundamental challenges in immunology. ACKNOWLEDGMENTS We thank many colleagues for helpful discussions and sharing unpublished data. We apologize to those colleagues whose studies were not cited owing to space limitations. We thank pediatricians and internists worldwide, without whom the search for mycobacterial susceptibility genes would be impossible. We thank students and postdoctoral fellows in the laboratory for their precious collaboration. We thank Martine Tridde-Mazloum for her trust and support. This work was supported by grants from the Action Blanche Jeunes Chercheurs, Fondation BNP-Paribas, Fondation pour la Recherche M´edicale, Fondation Schlumberger, Programme National de Recherche Fondamentale en Microbiologie et Maladies Infectieuses et Parasitaires, and the Sequella Foundation. Visit the Annual Reviews home page at www.annualreviews.org

LITERATURE CITED 1. Embley TM, Stackebrandt E. 1994. The molecular phylogeny and systematics of the actinomycetes. Annu. Rev. Microbiol. 48:257–89 2. Kubica GP, Wayne LG. 1984. The Mycobacteria: A Sourcebook. New York: Dekker. 2 Vol. 1553 pp. 3. Hatfull GF, Jacobs WR. 2000. Molecular Genetics of Mycobacteria. Washington, DC: ASM Press. 363 pp. 4. Rastogi N, Legrand E, Sola C. 2001. The mycobacteria: an introduction to nomenclature and pathogenesis. Rev. Sci. Tech. 20:21–54

5. Rogall T, Wolters J, Flohr T, Bottger EC. 1990. Towards a phylogeny and definition of species at the molecular level within the genus Mycobacterium. Int. J. Syst. Bacteriol. 40:323–30 6. Kim BJ, Lee SH, Lyu MA, Kim SJ, Bai GH, Chae GT, Kim EC, Cha CY, Kook YH. 1999. Identification of mycobacterial species by comparative sequence analysis of the RNA polymerase gene (rpoB). J. Clin. Microbiol. 37:1714–20 7. Ringuet H, Akoua-Koffi C, Honore S, Varnerot A, Vincent V, Berche P, Gaillard JL, Pierre-Audigier C. 1999. hsp65

2 Feb 2002

11:3

608

8.

Annu. Rev. Immunol. 2002.20:581-620. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

9.

10.

11.

12.

13.

14.

15. 16.

17.

18.

AR

AR152-19.tex

CASANOVA

¥

AR152-19.SGM

LaTeX2e(2001/05/10)

P1: GJC

ABEL

sequencing for identification of rapidly growing mycobacteria. J. Clin. Microbiol. 37:852–57 Behr MA, Wilson MA, Gill WP, Salamon H, Schoolnik GK, Rane S, Small PM. 1999. Comparative genomics of BCG vaccines by whole-genome DNA microarray. Science 284:1520–23 Kato-Maeda M, Bifani PJ, Kreiswirth BN, Small PM. 2001. The nature and consequence of genetic variability within Mycobacterium tuberculosis. J. Clin. Invest. 107:533–37 Falkinham JO. 1996. Epidemiology of infection by nontuberculous mycobacteria. Clin. Microbiol. Rev. 9:177–215 Wolinsky E, Rynearson TK. 1968. Mycobacteria in soil and their relation to disease-associated strains. Am. Rev. Respir. Dis. 97:1032–37 Collins CH, Grange JM, Yates MD. 1984. Mycobacteria in water. J. Appl. Bacteriol. 57:193–211 Falkinham JO 3rd, Norton CD, LeChevallier MW. 2001. Factors influencing numbers of Mycobacterium avium, Mycobacterium intracellulare, and other Mycobacteria in drinking water distribution systems. Appl. Environ. Microbiol. 67:1225–31 Bloom BR. 1994. Tuberculosis: Pathogenesis, Protection, and Control. Washington, DC: ASM Press. 637 pp. Rom WN, Garay SM. 1996. Tuberculosis. Boston: Little Brown. 1002 pp. Collins CH. 2000. The bovine tubercle bacillus. Br. J. Biomed. Sci. 57:234– 40 Stienstra Y, van Der Graaf WT, te Meerman GJ, The TH, de Leij LF, van der Werf TS. 2001. Susceptibility to development of Mycobacterium ulcerans disease: review of possible risk factors. Trop. Med. Int. Health 6:554–62 van der Werf TS, van der Graaf WT, Tappero JW, Asiedu K. 1999. Mycobacterium ulcerans infection. Lancet 354:1013–18

19. Hastings RC. 1994. Leprosy. Edinburgh/ New York: Churchill Livingstone. xiv, 470 pp 20. Jacobson RR, Krahenbuhl JL. 1999. Leprosy. Lancet 353:655–60 21. Bercovier H, Vincent V. 2001. Mycobacterial infections in domestic and wild animals due to Mycobacterium marinum, M. fortuitum, M. chelonae, M. porcinum, M. farcinogenes, M. smegmatis, M. scrofulaceum, M. xenopi, M. kansasii, M. simiae and M. genavense. Rev. Sci. Tech. 20:265– 90 22. Lander ES, Linton LM, Birren B, Nusbaum C, Zody MC, et al. 2001. Initial sequencing and analysis of the human genome. Nature 409:860–921 23. Venter JC, Adams MD, Myers EW, Li PW, Mural RJ, Sutton GG, Smith HO, Yandell M, Evans CA, Holt RA, et al. 2001. The sequence of the human genome. Science 291:1304–51 24. Cole ST, Brosch R, Parkhill J, Garnier T, Churcher C, Harris D, Gordon SV, Eiglmeier K, Gas S, Barry CE 3rd, et al. 1998. Deciphering the biology of Mycobacterium tuberculosis from the complete genome sequence. Nature 393:537– 44 25. Cole ST, Eiglmeier K, Parkhill J, James KD, Thomson NR, Wheeler PR, Honore N, Garnier T, Churcher C, Harris D, et al. 2001. Massive gene decay in the leprosy bacillus. Nature 409:1007– 11 26. WHO. 2000. Leprosy-global situation. Wkly. Epidemiol. Rec. 75:226–31 27. Dye C, Scheele S, Dolin P, Pathania V, Raviglione MC. 1999. Consensus statement. Global burden of tuberculosis: estimated incidence, prevalence, and mortality by country. WHO Global Surveillance and Monitoring Project. JAMA 282:677– 86 28. WHO. 2000. Buruli ulcer. Wkly. Epidemiol. Rec. 75:106–8 29. Lotte A, Wasz-Hockert O, Poisson N, Dumitrescu N, Verron M, Couvet E.

2 Feb 2002

11:3

AR

AR152-19.tex

AR152-19.SGM

LaTeX2e(2001/05/10)

P1: GJC

HUMAN GENETICS OF MYCOBACTERIA

Annu. Rev. Immunol. 2002.20:581-620. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

30.

31.

32.

33.

34.

35.

35a.

36.

37.

38.

39.

40.

1984. A bibliography of the complications of BCG vaccination. A comprehensive list of the world literature since the introduction of BCG up to July 1982, supplemented by over 100 personal communications. Adv. Tuberc. Res. 21:194–245 Casadevall A, Pirofski LA. 2000. Hostpathogen interactions: basic concepts of microbial commensalism, colonization, infection, and disease. Infect. Immun. 68:6511–18 Bloom BR, Small PM. 1998. The evolving relation between humans and Mycobacterium tuberculosis. N. Engl. J. Med. 338:677–78 McMurray DN. 2001. Disease model: pulmonary tuberculosis. Trends Mol. Med. 7:135–37 Stenger S, Modlin RL. 1999. T cell mediated immunity to Mycobacterium tuberculosis. Curr. Opin. Microbiol. 2:89–93 Flynn JL, Ernst JD. 2000. Immune responses in tuberculosis. Curr. Opin. Immunol. 12:432–36 Flynn JL, Chan J. 2001. Immunology of tuberculosis. Annu. Rev. Immunol. 19:93– 129 Pierce C, Dubos RJ, Middlebrook G. 1947. Infection of mice with mammalian tubercle bacilli grown in tween-albumin liquid medium. J. Exp. Med. 86:159–73 Skamene E, Gros P, Forget A, Kongshavn PA, St Charles C, Taylor BA. 1982. Genetic regulation of resistance to intracellular pathogens. Nature 297:506–9 Vidal S, Malo D, Vogan K, Skamene E, Gros P. 1993. Natural resistance to infection with intracellular parasites: isolation of a candidate for Bcg. Cell 73:469–85 Gruenheid S, Gros P. 2000. Genetic susceptibility to intracellular infections: Nramp1, macrophage function and divalent cations transport. Curr. Opin. Microbiol. 3:43–8 Buu N, Sanchez F, Schurr E. 2000. The Bcg host-resistance gene. Clin. Infect. Dis. 31(Suppl 3):S81–85 Kramnik I, Dietrich WF, Demant P,

41.

42. 43.

44.

45.

46.

47.

48.

49.

50.

609

Bloom BR. 2000. Genetic control of resistance to experimental infection with virulent Mycobacterium tuberculosis. Proc. Natl. Acad. Sci. USA 97:8560–65 Grange JM. 1996. Mycobacteria and Human Disease. London/New York: Arnold/ Oxford Univ. Press. 230 pp. Holland SM. 2001. Nontuberculous mycobacteria. Am. J. Med. Sci. 321:49–55 Koscielnak E, Naumann L, Casanova JL, Ottenhoff THM. 2001. Disseminated Mycobacterium peregrinum infection in a child with complete IFNγ R1 deficiency. Submitted Pierre-Audigier C, Jouanguy E, Lamhamedi S, Altare F, Rauzier J, Vincent V, Canioni D, Emile JF, Fischer A, Blanche S, Gaillard JL, Casanova JL. 1997. Fatal disseminated Mycobacterium smegmatis infection in a child with inherited interferon gamma receptor deficiency. Clin. Infect. Dis. 24:982–84 Wallace RJ Jr, Brown BA, Griffith DE. 1998. Nosocomial outbreaks/pseudooutbreaks caused by nontuberculous mycobacteria. Annu. Rev. Microbiol. 52: 453–90 Reichenbach J, Rosenzweig S, D¨offinger R, Dupuis S, Holland SM, Casanova JL. 2001. Mycobacterial diseases in primary immunodeficiencies. Curr. Opin. Clin. Immunol. Allergy 1:503–11 Oliver A, Maiz L, Canton R, Escobar H, Baquero F, Gomez-Mampaso E. 2001. Nontuberculous mycobacteria in patients with cystic fibrosis. Clin. Infect. Dis. 32:1298–303 WHO. 1999. Primary immunodeficiency diseases. Report of an IUIS Scientific Committee. International Union of Immunological Societies. Clin. Exp. Immunol. 118(Suppl 1):1–28 Fischer A. 2001. Primary immunodeficiency diseases: an experimental model for molecular medicine. Lancet 357: 1863–69 Ochs H, Smith CIE, Puck J. 2002. Primary Immunodeficiencies: A Molecular

2 Feb 2002

11:3

610

51.

Annu. Rev. Immunol. 2002.20:581-620. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

52.

53.

54.

55.

56.

57.

58.

AR

AR152-19.tex

CASANOVA

¥

AR152-19.SGM

LaTeX2e(2001/05/10)

P1: GJC

ABEL

and Genetic Approach. New York: Oxford Univ. Press Hugosson C, Harfi H. 1991. Disseminated BCG-osteomyelitis in congenital immunodeficiency. Pediatr. Radiol. 21: 384–85 Abramowsky C, Gonzalez B, Sorensen RU. 1993. Disseminated bacillus Calmette-Guerin infections in patients with primary immunodeficiencies. Am. J. Clin. Pathol. 100:52–56 Fisch P, Millner M, Muller SM, Wahn U, Friedrich W, Renz H. 1999. Expansion of gammadelta T cells in an infant with severe combined immunodeficiency syndrome after disseminated BCG infection and bone marrow transplantation. J. Allergy Clin. Immunol. 103:1218–19 Heyderman RS, Morgan G, Levinsky RJ, Strobel S. 1991. Successful bone marrow transplantation and treatment of BCG infection in two patients with severe combined immunodeficiency. Eur. J. Pediatr. 150:477–80 Minegishi M, Tsuchiya S, Imaizumi M, Yamaguchi Y, Goto Y, Tamura M, Konno T, Tada K. 1985. Successful transplantation of soy bean agglutinin-fractionated, histoincompatible, maternal marrow in a patient with severe combined immunodeficiency and BCG infection. Eur. J. Pediatr. 143:291–94 Gonzalez B, Moreno S, Burdach R, Valenzuela MT, Henriquez A, Ramos MI, Sorensen RU. 1989. Clinical presentation of Bacillus Calmette-Guerin infections in patients with immunodeficiency syndromes. Pediatr. Infect. Dis. J. 8:201– 6 Romanus V, Fasth A, Tordai P, Wiholm BE. 1993. Adverse reactions in healthy and immunocompromised children under six years of age vaccinated with the Danish BCG vaccine, strain Copenhagen 1331: implications for the vaccination policy in Sweden. Acta Paediatr. 82:1043–52 Stephan JL, Vlekova V, Le Deist F,

59.

60.

61.

62.

63.

64.

65.

Blanche S, Donadieu J, De Saint-Basile G, Durandy A, Griscelli C, Fischer A. 1993. Severe combined immunodeficiency: a retrospective single-center study of clinical presentation and outcome in 117 patients. J. Pediatr. 123:564–72 Kiehn TE, Edwards FF, Brannon P, Tsang AY, Maio M, Gold JW, Whimbey E, Wong B, McClatchy JK, Armstrong D. 1985. Infections caused by Mycobacterium avium complex in immunocompromised patients: diagnosis by blood culture and fecal examination, antimicrobial susceptibility tests, and morphological and seroagglutination characteristics. J. Clin. Microbiol. 21:168–73 Parent LJ, Salam MM, Appelbaum PC, Dossett JH. 1995. Disseminated Mycobacterium marinum infection and bacteremia in a child with severe combined immunodeficiency. Clin. Infect. Dis. 21:1325–27 Pasic S, Lilic D, Pejnovic N, Vojvodic D, Simic R, Abinun M. 1998. Disseminated Bacillus Calmette-Guerin infection in a girl with hyperimmunoglobulin E syndrome. Acta Paediatr. 87:702–4 Grimbacher B, Holland SM, Gallin JI, Greenberg F, Hill SC, Malech HL, Miller JA, O’Connell AC, Puck JM. 1999. Hyper-IgE syndrome with recurrent infections—an autosomal dominant multisystem disorder. N. Engl. J. Med. 340:692–702 Borges WG, Augustine NH, Hill HR. 2000. Defective interleukin-12/interferon-gamma pathway in patients with hyperimmunoglobulinemia E syndrome. J. Pediatr. 136:176–80 Verronen P. 1974. Case report. Presumed disseminated BCG in a boy with chronic granulomatous disease of childhood. Acta Paediatr. Scand. 63:627–30 Jacob CM, Pastorino AC, Azevedo AM, Marques HH, Sato HK, Ferrazole L, Aquino MZ, Sakane PT, Grumach AS. 1996. Mycobacterium bovis dissemination (BCG strain) among immunodeficient

2 Feb 2002

11:3

AR

AR152-19.tex

AR152-19.SGM

LaTeX2e(2001/05/10)

P1: GJC

HUMAN GENETICS OF MYCOBACTERIA

66.

Annu. Rev. Immunol. 2002.20:581-620. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

67.

68.

69.

70.

71.

72.

73.

74.

75.

Brazilian infants. J. Invest. Allergol. Clin. Immunol. 6:202–6 Keller E. 1976. Clinical picture of a generalized BCG vaccination tuberculosis as an expression of a progressive septic granulomatosis (author’s transl). Z. Erkr. Atmungsorgane 146:134–39 Smith PA, Wittenberg DF. 1984. Disseminated BCG infection in a child with chronic granulomatous disease. A case report. S. Afr. Med. J. 65:821–22 Kobayashi Y, Komazawa Y, Kobayashi M, Matsumoto T, Sakura N, Ishikawa K, Usui T. 1984. Presumed BCG infection in a boy with chronic granulomatous disease. A report of a case and a review of the literature. Clin. Pediatr. 23:586–89 Urban C, Becker H, Mutz I, Fritsch G. 1980. BCG-infection in chronic granulomatous disease (author’s transl). Klin. Padiatr. 192:13–18 Allen DM, Chng HH. 1993. Disseminated Mycobacterium flavescens in a probable case of chronic granulomatous disease. J. Infect. 26:83–86 Moskaluk CA, Pogrebniak HW, Pass HI, Gallin JI, Travis WD. 1994. Surgical pathology of the lung in chronic granulomatous disease. Am. J. Clin. Pathol. 102:684–91 Chusid MJ, Parrillo JE, Fauci AS. 1975. Chronic granulomatous disease. Diagnosis in a 27-year-old man with Mycobacterium fortuitum. JAMA 233:1295–96 Ohga S, Ikeuchi K, Kadoya R, Okada K, Miyazaki C, Suita S, Ueda K. 1997. Intrapulmonary Mycobacterium avium infection as the first manifestation of chronic granulomatous disease. J. Infect. 34:147– 50 Sitton JE, Reimund EL. 1992. Extramedullary hematopoiesis of the cranial dura and anhidrotic ectodermal dysplasia. Neuropediatrics 23:108–10 Doffinger R, Smahi A, Bessia C, Geissmann F, Feinberg J, Durandy A, Bodemer C, Kenwrick S, Dupuis-Girod S, Blanche S, Wood P, Rabia SH, Headon

76.

77.

78.

79.

80.

81.

82.

611

DJ, Overbeek PA, Le Deist F, Holland SM, Belani K, Kumararatne DS, Fischer A, Shapiro R, Conley ME, Reimund E, Kalhoff H, Abinun M, Munnich A, Israel A, Courtois G, Casanova JL. 2001. X-linked anhidrotic ectodermal dysplasia with immunodeficiency is caused by impaired NF-kappaB signaling. Nat. Genet. 27:277–85 Aradhya S, Courtois G, Rajkovic A, Lewis R, Levy M, Israel A, Nelson D. 2001. Atypical forms of incontinentia pigmenti in male individuals result from mutations of a cytosine tract in exon 10 of NEMO (IKK-gamma). Am. J. Hum. Genet. 68:765–71 Dupuis-Girod S, Corradini N, Doeffinger R, Smahi A, Courtois G, Israel A, Munnich A, Fischer A, Casanova JL, Bodemer C. 2002. Incontinentia pigmenti and anhydrotic ectodermal dysplasia with immunodeficiency, lymphedema and osteopetrosis in a single kindred. Submitted Mansour S, Woffendin H, Mitton S, Jeffery I, Jakins T, Kenwrick S, Murday VA. 2001. Incontinentia pigmenti in a surviving male is accompanied by hypohidrotic ectodermal dysplasia and recurrent infection. Am. J. Med. Genet. 99:172–77 Brooks EG, Klimpel GR, Vaidya SE, Keeney SE, Raimer S, Goldman AS, Schmalstieg FC. 1994. Thymic hypoplasia and T-cell deficiency in ectodermal dysplasia: case report and review of the literature. Clin. Immunol. Immunopathol. 71:44–52 Levin M, Newport MJ, D’Souza S, Kalabalikis P, Brown IN, Lenicker HM, Agius PV, Davies EG, Thrasher A, Klein N, Blackwell J. 1995. Familial disseminated atypical mycobacterial infection in childhood: a human mycobacterial susceptibility gene? Lancet 345:79–83 Casanova JL, Jouanguy E, Lamhamedi S, Blanche S, Fischer A. 1995. Immunological conditions of children with BCG disseminated infection. Lancet 346:581 Casanova JL, Blanche S, Emile JF,

2 Feb 2002

11:3

612

Annu. Rev. Immunol. 2002.20:581-620. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

83.

84.

85.

86.

87.

88.

AR

AR152-19.tex

CASANOVA

¥

AR152-19.SGM

LaTeX2e(2001/05/10)

P1: GJC

ABEL

Jouanguy E, Lamhamedi S, Altare F, Stephan JL, Bernaudin F, Bordigoni P, Turck D, Lachaux A, Albertini M, Bourrillon A, Dommergues JP, Pocidalo MA, Le Deist F, Gaillard JL, Griscelli C, Fischer A. 1996. Idiopathic disseminated bacillus Calmette-Guerin infection: a French national retrospective study. Pediatrics 98:774–78 Frucht DM, Holland SM. 1996. Defective monocyte costimulation for IFNgamma production in familial disseminated Mycobacterium avium complex infection: abnormal IL-12 regulation. J. Immunol. 157:411–16 McKusick VA. 1998. Mendelian Inheritance in Man. Catalogs of Human Genes and Genetic Disorders. Baltimore: Johns Hopkins Univ. Press Frucht DM, Sandberg DI, Brown MR, Gerstberger SM, Holland SM. 1999. IL-12-independent costimulation pathways for interferon-γ production in familial disseminated Mycobacterium avium complex infection. Clin. Immunol. 91:234–41 Jouanguy E, Lamhamedi-Cherradi S, Lammas D, Dorman SE, Fondan`eche M-C, Dupuis S, D¨offinger R, Altare F, Emile J-F, Girdelstone J, Ducoulombier H, Edgar D, Clarke J, Oxelius VA, Brai M, Novelli V, Heyne K, Fischer A, Holland SM, Kumararatne DS, Schreiber RD, Casanova J-L. 1999. A human IFNGR1 small deletion hotspot associated with dominant susceptibility to mycobacterial infection. Nat. Genet. 21:370– 78 Dupuis S, Dargemont C, Fieschi C, Thomassin N, Rosenzweig S, Harris J, Holland SM, Schreiber RD, Casanova JL. 2001. Impairment of mycobacterial but not viral immunity by a germline human STAT1 mutation. Science 293:300–3 Emile JF, Patey N, Altare F, Lamhamedi S, Jouanguy E, Boman F, Quillard J, Lecomte-Houcke M, Verola O, Mousnier JF, Dijoud F, Blanche S, Fischer A,

88a.

89.

90.

91.

92.

93.

94.

95.

Brousse N, Casanova JL. 1997. Correlation of granuloma structure with clinical outcome defines two types of idiopathic disseminated BCG infection. J. Pathol. 181:25–30 Nau GJ, Chupp GL, Emile JF, Jouanguy E, Beiman JS, Casanova JL, Young RA. 2001. Osteopontin expression correlates with clinical outcome in patients with mycobacterial infection. Am. J. Pathol. 157:37–42 Lammas DA, Casanova JL, Kumararatne DS. 2000. Clinical consequences of defects in the IL-12-dependent interferongamma (IFN-gamma) pathway. Clin. Exp. Immunol. 121:417–25 Dorman SE, Holland SM. 2000. Interferon-gamma and interleukin-12 pathway defects and human disease. Cytokine Growth Factor Rev. 11:321–33 Picard C, Baud O, Fieschi C, Casanova JL. 2000. Diagnosis and management of inheritable disorders of IFNγ -mediated immunity. Immunol. Allergy Clin. North Am. 20:65–76 Remus N, Reichenbach J, Picard C, Rietschel C, Wood P, Lammas D, Kumararatne DS, Casanova JL. 2001. Impaired interferon gamma-mediated immunity and susceptibility to mycobacterial infection in childhood. Pediatr. Res. 50:8–13 Newport M, Holland SM, Levin M, Casanova JL. 2002. Inherited disorders of the IL-12-IFNγ axis. In Primary Immunodeficiencies: A Molecular and Genetic Approach, ed. H Ochs, CIE Smith, J Puck. New York: Oxford Univ. Press Newport MJ, Huxley CM, Huston S, Hawrylowicz CM, Oostra BA, Williamson R, Levin M. 1996. A mutation in the interferon-gamma-receptor gene and susceptibility to mycobacterial infection. N. Engl. J. Med. 335:1941–49 Jouanguy E, Altare F, Lamhamedi S, Revy P, Emile JF, Newport M, Levin M, Mlanche S, Seboun E, Fischer A, Casanova JL. 1996. Interferon-gamma-receptor deficiency in an infant with fatal bacille

2 Feb 2002

11:3

AR

AR152-19.tex

AR152-19.SGM

LaTeX2e(2001/05/10)

P1: GJC

HUMAN GENETICS OF MYCOBACTERIA

Annu. Rev. Immunol. 2002.20:581-620. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

96.

97.

98.

99.

99a.

99b.

100.

Calmette-Guerin infection. N. Engl. J. Med. 335:1956–61 Altare F, Jouanguy E, LamhamediCherradi S, Fondaneche MC, Fizame C, Ribierre F, Merlin G, Dembic Z, Schreiber R, Lisowska-Grospierre B, Fischer A, Seboun E, Casanova JL. 1998. A causative relationship between mutant IFNgR1 alleles and impaired cellular response to IFNγ in a compound heterozygous child. Am. J. Hum. Genet. 62:723–26 Roesler J, Kofink B, Wendisch J, Heyden S, Paul D, Lehmann R, Friedrich W, Casanova JL, Leupold W, Gahr M, R¨osen-Wolff A. 1999. Recurrent mycobacterial and listeria infections in a child with interferon γ receptor deficiency: mutational analysis and evaluation of therapeutic options. Exp. Haematol. 27:1368–74 Holland SM, Dorman SE, Kwon A, Pitha-Rowe IF, Frucht DM, Gerstberger SM, Noel GJ, Vesterhus P, Brown MR, Fleisher TA. 1998. Abnormal regulation of interferon gamma, interleukin 12, and tumor necrosis factor alpha in interferon gamma receptor 1 deficiency. J. Infect. Dis. 178:1095–104 Vesterhus P, Holland SM, Abrahamsen TG, Bjerknes R. 1998. Familial disseminated infection due to atypical mycobacteria with childhood onset. Clin. Infect. Dis. 27:822–25 Cunningham JA, Kollner JD, Bridge PJ, Trevenen CL, McLeod DR, Davies HD. 2000. Discriminated bacille CalmetteGu´erin infection in an infant with a novel deletion in the interferon-gamma receptor gene. Int. J. Tuberc. Lung Dis. 4:791–94 Rosenzweig S, Dorman SE, Roesler J, Zelazko M, Holland SM. 2002. A novel autosomal recessive mutation defines a second mutational hotspot in the interferon gamma receptor 1 (IFNgammaR1) chain. Clin. Immunol. In press Dorman SE, Holland SM. 1998. Mutation in the signal-transducing chain of the interferon-gamma receptor and suscepti-

101.

102.

103.

104.

105.

106.

107.

613

bility to mycobacterial infection. J. Clin. Invest. 101:2364–69 Jouanguy E, Pallier A, Dupuis S, D¨offinger R, Fondan`eche MC, LamhamediCherradi S, Altare F, Emile JF, Lutz P, Bordigoni P, Cokugras H, Akcakaya N, Landman-Parker J, Donnadieu J, Camcioglu Y, Casanova JL. 2000. In a novel form of complete IFNγ R1 deficiency, cell-surface receptors fail to bind IFNγ . J. Clin. Invest. 105:1429–36 Allende LM, Lopez-Goyanes A, PazArtal E, Corell A, Garcia-Perez MA, Varela P, Scarpellini A, Negreira S, Palenque E, Arnaiz-Villena A. 2001. A point mutation in a domain of gamma interferon receptor 1 provokes severe immunodeficiency. Clin. Diagn. Lab. Immunol. 8:133–37 Fieschi C, Dupuis S, Picard C, Smith CI, Holland SM, Casanova JL. 2001. High levels of interferon gamma in the plasma of children with complete interferon gamma receptor deficiency. Pediatrics 107:E48 Uzel G, Premkumar A, Malech HL, Holland SM. 2000. Respiratory syncytial virus infection in patients with phagocyte defects. Pediatrics 106:835–37 Dorman SE, Uzel G, Roesler J, Bradley JS, Bastian J, Billman G, King S, Filie A, Schermerhorn PA, Holland SM. 1999. Viral infections in interferon-γ receptor deficiency. J. Pediatr. 135:640–43 Jouanguy E, Lamhamedi-Cherradi S, Altare F, Fondaneche MC, Tuerlinckx D, Blanche S, Emile JF, Gaillard JL, Schreiber R, Levin M, Fischer A, Hivroz C, Casanova JL. 1997. Partial interferon-gamma receptor 1 deficiency in a child with tuberculoid bacillus CalmetteGuerin infection and a sibling with clinical tuberculosis. J. Clin. Invest. 100:2658– 64 D¨offinger R, Jouanguy E, Dupuis S, Fondan`eche MC, St´ephan JL, Emile JF, Lamhamedi S, Altare F, Pallier A, Barcenas-Morales G, Krause C, Pestka S,

2 Feb 2002

11:3

614

Annu. Rev. Immunol. 2002.20:581-620. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

108.

109.

109a.

109b.

109c.

110.

AR

AR152-19.tex

CASANOVA

¥

AR152-19.SGM

LaTeX2e(2001/05/10)

P1: GJC

ABEL

Schreiber RD, Novelli F, Casanova JL. 2000. Partial interferon gamma receptor signalling chain deficiency in a patient with bacille Calmette-Gu´erin and Mycobacterium abscessus infection. J. Infect. Dis. 181:379–84 Dupuis S, Doffinger R, Picard C, Fieschi C, Altare F, Jouanguy E, Abel L, Casanova JL. 2000. Human interferongamma-mediated immunity is a genetically controlled continuous trait that determines the outcome of mycobacterial invasion. Immunol. Rev. 178:129–37 Villella A, Picard C, Jouanguy E, Dupuis S, Popko S, Abughali N, Meyerson H, Casanova JL, Hostoffer RW. 2001. Recurrent Mycobacterium avium osteomyelitis associated with a novel dominant interferon gamma receptor mutation. Pediatrics 107:E47 Scardamaglia L, Tam WY, Nikpom M, Harrison LC, Street AC, Kay TWH. 2002. Disseminated non-tuberculous mycobacterial infection due to a novel interferon gamma receptor mutation. Submitted Sasaki Y, Nomura A, Kusuhara K, Takada H, Ahmed S, Takahata Y, Obinata K, Hamada K, Hara T. 2002. Genetic basis of patients with Bacille Calmette-Guerin osteomyelitis in Japan: identification of IFNgammaR1 defciency as a predominant type. J. Infect. Dis. In press Arend SM, Janssen R, Gosen JJ, Waanders H, de Boer T, Ottenhoff THM, van Dissel JT. 2001. Multifocal osteomyelitis caused by nontuberculous mycobacteria in patients with a genetic defect of the interferon-γ receptor. Netherlands J. Med. 59:140–141 Altare F, Lammas D, Revy P, Jouanguy E, D¨offinger R, Lamhamedi S, Drysdale P, Tollner D, Girdlestone J, Darbyshire P, Wadhwa M, Dockrel H, Salmon M, Fischer A, Durandy A, Casanova JL, Kumararatne D. 1998. Inherited interleukin 12 deficiency in a

111.

112.

113.

114.

115.

116.

116a.

child with bacille Calmette-Gu´erin and Salmonella enteritidis disseminated infection. J. Clin. Invest. 102:2035–40 Picard C, Fieschi C, Altare F, et al. 2002. Inherited IL-12 deficiency: IL12B genotype and clinical phenotype of 13 patients from five kindreds. Am. J. Hum. Genet. In press Altare F, Durandy A, Lammas D, Emile JF, Lamhamedi S, Le Deist F, Drysdale P, Jouanguy E, D¨offinger R, Bernaudin F, Jeppsson O, Gollob JA, Meinl E, Segal AW, Fischer A, Kumararatne D, Casanova JL. 1998. Impairment of mycobacterial immunity in human interleukin-12 receptor deficiency. Science 280:1432–35 de Jong R, Altare F, Haagen IA, Elferink DG, Boer T, van Breda Vriesman PJ, Kabel PJ, Draaisma JM, van Dissel JT, Kroon FP, Casanova JL, Ottenhoff TH. 1998. Severe mycobacterial and Salmonella infections in interleukin12 receptor-deficient patients. Science 280:1435–38 Aksu G, Tirpan C, Cavusoglu C, Soydan S, Altare F, Casanova JL, Kutukculer N. 2001. Mycobacterium fortuitum-chelonae complex infection in a child with complete interleukin-12 receptor beta 1 deficiency. Pediatr. Infect. Dis. J. 20:551–53 Altare F, Ensser A, Breiman A, Reichenbach J, Baghdadi JE, Fischer A, Emile JF, Gaillard JL, Meinl E, Casanova JL. 2001. Interleukin-12 receptor beta1 deficiency in a patient with abdominal tuberculosis. J. Infect. Dis. 184:231– 36 Sakai T, Matsuoka M, Aoki M, Nosaka K, Mitsuya H. 2001. Missense mutation of the interleukin-12 receptor beta1 chain-encoding gene is associated with impaired immunity against Mycobacterium avium complex infection. Blood 97:2688–94 Losana G, Rigamouti L, Borghi I, Ariotti S, Jouanguy E, Altare F, Forni G,

2 Feb 2002

11:3

AR

AR152-19.tex

AR152-19.SGM

LaTeX2e(2001/05/10)

P1: GJC

HUMAN GENETICS OF MYCOBACTERIA

Annu. Rev. Immunol. 2002.20:581-620. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

116b.

117.

118.

119.

120.

121.

122.

Casanova JL, Novelli F. 2002. Requirement for both IL-12 and IFN-gamma signaling pathways in optimal IFNgamma production by human T cells. Submitted Ward MR, Kagan R, Engert JC, Lepage P, Raby B, Remus N, Kovesi T, Casanova JL, Hudson TJ. 2002. Genetic analysis of host susceptibility in patients with unusual mycobacterial infections. Submitted Gollob JA, Veenstra KG, Jyonouchi H, Kelly AM, Ferrieri P, Panka DJ, Altare F, Fieschi C, Casanova JL, Frank DA, Mier JW. 2000. Impairment of STAT activation by IL-12 in a patient with atypical mycobacterial and staphylococcal infections. J. Immunol. 165:4120–26 Verhagen CE, de Boer T, Smits HH, Verreck FA, Wierenga EA, Kurimoto M, Lammas DA, Kumararatne DS, Sanal O, Kroon FP, van Dissel JT, Sinigaglia F, Ottenhoff TH. 2000. Residual type 1 immunity in patients genetically deficient for interleukin 12 receptor beta1 (IL-12Rbeta1): evidence for an IL12Rbeta1-independent pathway of IL-12 responsiveness in human T cells. J. Exp. Med. 192:517–28 D¨offinger R, Jouanguy E, Altare F, Wood P, Shirakawa T, Novelli F, Lammas D, Kumararatne D, Casanova J-L. 1999. Inheritable defects in IL-12and IFNγ -mediated immunity and the TH1/TH2 paradigm in man. Allergy 54:409–12 Jouanguy E, D¨offinger R, Dupuis S, Pallier A, Altare F, Casanova JL. 1999. IL-12 and IFN-γ in host defense against mycobacteria and salmonella in mice and men. Curr. Opin. Immunol. 11:346– 51 Jouanguy E, Altare F, LamhamediCherradi S, Casanova JL. 1997. Infections in IFNGR-1-deficient children. J. Interferon. Cytokine Res. 17:583–87 Bloom BR, Murray CJ. 1992. Tuberculosis: commentary on a reemergent killer. Science 257:1055–64

615

123. Marquet S, Schurr E. 2001. Genetics of susceptibility to infectious diseases: tuberculosis and leprosy as examples. Drug Metab. Dispos. 29:479–83 124. Stead WW. 1992. Genetics and resistance to tuberculosis. Could resistance be enhanced by genetic engineering? Ann. Intern. Med. 116:937–41 125. Stead WW. 1997. The origin and erratic global spread of tuberculosis. How the past explains the present and is the key to the future. Clin. Chest Med. 18:65– 77 126. Stead WW, Senner JW, Reddick WT, Lofgren JP. 1990. Racial differences in susceptibility to infection by Mycobacterium tuberculosis. N. Engl. J. Med. 322:422–27 127. Comstock GW. 1978. Tuberculosis in twins: a re-analysis of the Prophit survey. Am. Rev. Respir. Dis. 117:621–24 128. Chakravartti MR, Vogel FA. 1973. A twin study of leprosy. Top. Hum. Genet. 1:1–123 129. Abel L, Lap VD, Oberti J, Thuc NV, Cua VV, Guilloud-Bataille M, Schurr E, Lagrange PH. 1995. Complex segregation analysis of leprosy in southern Vietnam. Genet. Epidemiol. 12:63–82 130. Abel L, Demenais F. 1988. Detection of major genes for susceptibility to leprosy and its subtypes in a Caribbean island: Desirade Island. Am. J. Hum. Genet. 42: 256–66 131. Deerojanawong J, Chang AB, Eng PA, Robertson CF, Kemp AS. 1997. Pulmonary diseases in children with severe combined immune deficiency and DiGeorge syndrome. Pediatr. Pulmonol. 24:324–30 132. Levy J, Espanol-Boren T, Thomas C, Fischer A, Tovo P, Bordigoni P, Resnick I, Fasth A, Baer M, Gomez L, Sanders EA, Tabone MD, Plantaz D, Etzioni A, Monafo V, Abinun M, Hammarstrom L, Abrahamsen T, Jones A, Finn A, Klemola T, DeVries E, Sanal O, Peitsch MC, Notarangelo LD. 1997. Clinical

2 Feb 2002

11:3

616

133.

Annu. Rev. Immunol. 2002.20:581-620. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

134.

135.

136.

137.

138.

139.

140.

141.

AR

AR152-19.tex

CASANOVA

¥

AR152-19.SGM

LaTeX2e(2001/05/10)

P1: GJC

ABEL

spectrum of X-linked hyper-IgM syndrome. J. Pediatr. 131:47–54 Winkelstein JA, Marino MC, Johnston RB Jr, Boyle J, Curnutte J, Gallin JI, Malech HL, Holland SM, Ochs H, Quie P, Buckley RH, Foster CB, Chanock SJ, Dickler H. 2000. Chronic granulomatous disease. Report on a national registry of 368 patients. Medicine 79:155–69 Lau YL, Chan GC, Ha SY, Hui YF, Yuen KY. 1998. The role of phagocytic respiratory burst in host defense against Mycobacterium tuberculosis. Clin. Infect. Dis. 26:226–27 Frix CD 3rd, Bronson DM. 1986. Acute miliary tuberculosis in a child with anhidrotic ectodermal dysplasia. Pediatr. Dermatol. 3:464–67 Hein R, Peest D, Qaiyumi SA, Kolanus W, Deicher H, Schmidt RE. 1990. Granulomatous inflammation with combined immunodeficiency. Immun. Infekt. 18:48–50 Ozsahin H, Arredondo-Vega FX, Santisteban I, Fuhrer H, Tuchschmid P, Jochum W, Aguzzi A, Lederman HM, Fleischman A, Winkelstein JA, Seger RA, Hershfield MS. 1997. Adenosine deaminase deficiency in adults. Blood 89:2849–55 Abel L, Casanova JL. 2000. Genetic predisposition to clinical tuberculosis: bridging the gap between simple and complex inheritance. Am. J. Hum. Genet. 67:274–77 Abel L, Dessein AJ. 1998. Genetic epidemiology of infectious diseases in humans: design of population-based studies. Emerg. Infect. Dis. 4:593–603 Dib C, Faure S, Fizames C, Samson D, Drouot N, Vignal A, Millasseau P, Marc S, Hazan J, Seboun E, Lathrop M, Gyapay G, Morissette J, Weissenbach J. 1996. A comprehensive genetic map of the human genome based on 5,264 microsatellites. Nature 380:152–54 Sachidanandam R, Weissman D, Schmidt SC, Kakol JM, Stein LD, Marth

142.

143.

144.

145.

146.

147.

148.

G, Sherry S, Mullikin JC, Mortimore BJ, Willey DL, et al. 2001. A map of human genome sequence variation containing 1.42 million single nucleotide polymorphisms. Nature 409:928–33 Wang DG, Fan JB, Siao CJ, Berno A, Young P, Sapolsky R, Ghandour G, Perkins N, Winchester E, Spencer J, Kruglyak L, Stein L, Hsie L, Topaloglou T, Hubbell E, Robinson E, Mittmann M, Morris MS, Shen N, Kilburn D, Rioux J, Nusbaum C, Rozen S, Hudson TJ, Lipshutz R, Chee M, Lander ES. 1998. Large-scale identification, mapping, and genotyping of single-nucleotide polymorphisms in the human genome. Science 280:1077– 82 Lander ES, Schork NJ. 1994. Genetic dissection of complex traits. Science 265:2037–48 Abel L, Dessein AJ. 1997. The impact of host genetics on susceptibility to human infectious diseases. Curr. Opin. Immunol. 9:509–16 Bothamley GH, Beck JS, Schreuder GM, D’Amaro J, de Vries RR, Kardjito T, Ivanyi J. 1989. Association of tuberculosis and M. tuberculosis-specific antibody levels with HLA. J. Infect. Dis. 159:549–55 Brahmajothi V, Pitchappan RM, Kakkanaiah VN, Sashidhar M, Rajaram K, Ramu S, Palanimurugan K, Paramasivan CN, Prabhakar R. 1991. Association of pulmonary tuberculosis and HLA in south India. Tubercle 72:123–32 Rajalingam R, Mehra NK, Jain RC, Myneedu VP, Pande JN. 1996. Polymerase chain reaction–based sequencespecific oligonucleotide hybridization analysis of HLA class II antigens in pulmonary tuberculosis: relevance to chemotherapy and disease severity. J. Infect. Dis. 173:669–76 Singh SP, Mehra NK, Dingley HB, Pande JN, Vaidya MC. 1983. Human leukocyte antigen (HLA)-linked control

2 Feb 2002

11:3

AR

AR152-19.tex

AR152-19.SGM

LaTeX2e(2001/05/10)

P1: GJC

HUMAN GENETICS OF MYCOBACTERIA

Annu. Rev. Immunol. 2002.20:581-620. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

149.

150.

151.

152.

153.

154.

155.

of susceptibility to pulmonary tuberculosis and association with HLA-DR types. J. Infect. Dis. 148:676–81 Hawkins BR, Higgins DA, Chan SL, Lowrie DB, Mitchison DA, Girling DJ. 1988. HLA typing in the Hong Kong Chest Service/British Medical Research Council study of factors associated with the breakdown to active tuberculosis of inactive pulmonary lesions. Am. Rev. Respir. Dis. 138:1616–21 Cox RA, Downs M, Neimes RE, Ognibene AJ, Yamashita TS, Ellner JJ. 1988. Immunogenetic analysis of human tuberculosis. J. Infect. Dis. 158:1302–8 Sanjeevi CB, Narayanan PR, Prabakar R, Charles N, Thomas BE, Balasubramaniam R, Olerup O. 1992. No association or linkage with HLA-DR or -DQ genes in south Indians with pulmonary tuberculosis. Tuber. Lung Dis. 73:280– 84 Schreuder GM, Hurley CK, Marsh SG, Lau M, Maiers M, Kollman C, Noreen H. 1999. The HLA dictionary 1999: a summary of HLA-A, -B, -C, -DRB1/ 3/4/5, -DQB1 alleles and their association with serologically defined HLA-A, -B, -C, -DR and -DQ antigens. Tissue Antigens 54:409–37 Mehra NK, Rajalingam R, Mitra DK, Taneja V, Giphart MJ. 1995. Variants of HLA-DR2/DR51 group haplotypes and susceptibility to tuberculoid leprosy and pulmonary tuberculosis in Asian Indians. Int. J. Lepr. Other Mycobact. Dis. 63:241–48 Ravikumar M, Dheenadhayalan V, Rajaram K, Lakshmi SS, Kumaran PP, Paramasivan CN, Balakrishnan K, Pitchappan RM. 1999. Associations of HLA-DRB1, DQB1 and DPB1 alleles with pulmonary tuberculosis in south India. Tuber. Lung Dis. 79:309–17 Teran-Escandon D, Teran-Ortiz L, Camarena-Olvera A, Gonzalez-Avila G, Vaca-Marin MA, Granados J, Selman M. 1999. Human leukocyte antigen-

156.

157.

158.

159.

160.

161.

162.

163.

164.

617

associated susceptibility to pulmonary tuberculosis: molecular analysis of class II alleles by DNA amplification and oligonucleotide hybridization in Mexican patients. Chest 115:428–33 Goldfeld AE, Delgado JC, Thim S, Bozon MV, Uglialoro AM, Turbay D, Cohen C, Yunis EJ. 1998. Association of an HLA-DQ allele with clinical tuberculosis. JAMA 279:226–28 Ottenhoff TH, de Vries RR. 1987. HLA class II immune response and suppression genes in leprosy. Int. J. Lepr. Other Mycobact. Dis. 55:521–34 Meyer CG, May J, Stark K. 1998. Human leukocyte antigens in tuberculosis and leprosy. Trends Microbiol. 6:148–54 de Vries RR, Mehra NK, Vaidya MC, Gupte MD, Meera Khan P, Van Rood JJ. 1980. HLA-linked control of susceptibility to tuberculoid leprosy and association with HLA-DR types. Tissue Antigens 16:294–304 Dessoukey MW, el-Shiemy S, Sallam T. 1996. HLA and leprosy: segregation and linkage study. Int. J. Dermatol. 35:257– 64 Rani R, Fernandez-Vina MA, Zaheer SA, Beena KR, Stastny P. 1993. Study of HLA class II alleles by PCR oligotyping in leprosy patients from north India. Tissue Antigens 42:133–37 Zerva L, Cizman B, Mehra NK, Alahari SK, Murali R, Zmijewski CM, Kamoun M, Monos DS. 1996. Arginine at positions 13 or 70–71 in pocket 4 of HLADRB1 alleles is associated with susceptibility to tuberculoid leprosy. J. Exp. Med. 183:829–36 de Vries RR, Fat RF, Nijenhuis LE, van Rood JJ. 1976. HLA-linked genetic control of host response to Mycobacterium leprae. Lancet 2:1328–30 Fine PE, Wolf E, Pritchard J, Watson B, Bradley DJ, Festenstein H, Chacko CJ. 1979. HLA-linked genes and leprosy: a family study in Karigiri, South India. J. Infect. Dis. 140:152–61

2 Feb 2002

11:3

Annu. Rev. Immunol. 2002.20:581-620. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

618

AR

AR152-19.tex

CASANOVA

¥

AR152-19.SGM

LaTeX2e(2001/05/10)

P1: GJC

ABEL

165. van Eden W, Gonzalez NM, de Vries RR, Convit J, van Rood JJ. 1985. HLA-linked control of predisposition to lepromatous leprosy. J. Infect. Dis. 151: 9–14 166. Xu KY, de Vries RR, Fei HM, van Leeuwen A, Chen RB, Ye GY. 1985. HLA-linked control of predisposition to lepromatous leprosy. Int. J. Lepr. Other Mycobact. Dis. 53:56–63 167. Blackwell JM, Black GF, Peacock CS, Miller EN, Sibthorpe D, Gnananandha D, Shaw JJ, Silveira F, Lins-Lainson Z, Ramos F, Collins A, Shaw MA. 1997. Immunogenetics of leishmanial and mycobacterial infections: the Belem Family Study. Philos. Trans. R. Soc. London Ser. B 352:1331–45 168. Roy S, McGuire W, Mascie-Taylor CG, Saha B, Hazra SK, Hill AV, Kwiatkowski D. 1997. Tumor necrosis factor promoter polymorphism and susceptibility to lepromatous leprosy. J. Infect. Dis. 176:530–32 169. Spies T, Bresnahan M, Bahram S, Arnold D, Blanck G, Mellins E, Pious D, DeMars R. 1990. A gene in the human major histocompatibility complex class II region controlling the class I antigen presentation pathway. Nature 348:744– 47 170. Rajalingam R, Singal DP, Mehra NK. 1997. Transporter associated with antigen-processing (TAP) genes and susceptibility to tuberculoid leprosy and pulmonary tuberculosis. Tissue Antigens 49:168–72 171. Vidal SM, Malo D, Vogan K, Skamene E, Gros P. 1993. Natural resistance to infection with intracellular parasites: isolation of a candidate for Bcg. Cell 73:469– 85 172. Brown IN, Glynn AA, Plant J. 1982. Inbred mouse strain resistance to Mycobacterium lepraemurium follows the Ity/Lsh pattern. Immunology 47:149– 56 173. North RJ, LaCourse R, Ryan L, Gros P.

174.

175.

176.

177.

178.

179.

180.

1999. Consequence of Nramp1 deletion to Mycobacterium tuberculosis infection in mice. Infect. Immun. 67:5811–14 Jabado N, Jankowski A, Dougaparsad S, Picard V, Grinstein S, Gros P. 2000. Natural resistance to intracellular infections: natural resistance-associated macrophage protein 1 (Nramp1) functions as a pH-dependent manganese transporter at the phagosomal membrane. J. Exp. Med. 192:1237–48 Cellier M, Govoni G, Vidal S, Kwan T, Groulx N, Liu J, Sanchez F, Skamene E, Schurr E, Gros P. 1994. Human natural resistance-associated macrophage protein: cDNA cloning, chromosomal mapping, genomic organization, and tissue-specific expression. J. Exp. Med. 180:1741–52 Bellamy R, Ruwende C, Corrah T, McAdam KP, Whittle HC, Hill AV. 1998. Variations in the NRAMP1 gene and susceptibility to tuberculosis in West Africans. N. Engl. J. Med. 338:640–44 Cervino AC, Lakiss S, Sow O, Hill AV. 2000. Allelic association between the NRAMP1 gene and susceptibility to tuberculosis in Guinea-Conakry. Ann. Hum. Genet. 64:507–12 Ryu S, Park YK, Bai GH, Kim SJ, Park SN, Kang S. 2000. 30 UTR polymorphisms in the NRAMP1 gene are associated with susceptibility to tuberculosis in Koreans. Int. J. Tuberc. Lung Dis. 4:577–80 Gao PS, Fujishima S, Mao XQ, Remus N, Kanda M, Enomoto T, Dake Y, Bottini N, Tabuchi M, Hasegawa N, Yamaguchi K, Tiemessen C, Hopkin JM, Shirakawa T, Kishi F. 2000. Genetic variants of NRAMP1 and active tuberculosis in Japanese populations. Clin. Genet. 58:74–76 Greenwood CM, Fujiwara TM, Boothroyd LJ, Miller MA, Frappier D, Fanning EA, Schurr E, Morgan K. 2000. Linkage of tuberculosis to chromosome 2q35 loci, including NRAMP1, in a large

2 Feb 2002

11:3

AR

AR152-19.tex

AR152-19.SGM

LaTeX2e(2001/05/10)

P1: GJC

HUMAN GENETICS OF MYCOBACTERIA

181.

Annu. Rev. Immunol. 2002.20:581-620. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

182.

183.

184.

185.

186.

187.

aboriginal Canadian family. Am. J. Hum. Genet. 67:405–16 Abel L, Sanchez FO, Oberti J, Thuc NV, Hoa LV, Lap VD, Skamene E, Lagrange PH, Schurr E. 1998. Susceptibility to leprosy is linked to the human NRAMP1 gene. J. Infect. Dis. 177:133–45 Shaw MA, Atkinson S, Dockrell H, Hussain R, Lins-Lainson Z, Shaw J, Ramos F, Silveira F, Mehdi SQ, Kaukab F, Khaliq S, Chiang T, Blackwell J. 1993. An RFLP map for 2q33–q37 from multicase mycobacterial and leishmanial disease families: no evidence for an Lsh/Ity/Bcg gene homologue influencing susceptibility to leprosy. Ann. Hum. Genet. 57:251–71 Levee G, Liu J, Gicquel B, Chanteau S, Schurr E. 1994. Genetic control of susceptibility to leprosy in French Polynesia: no evidence for linkage with markers on telomeric human chromosome 2. Int. J. Lepr. Other Mycobact. Dis. 62:499– 511 Alcais A, Sanchez FO, Thuc NV, Lap VD, Oberti J, Lagrange PH, Schurr E, Abel L. 2000. Granulomatous reaction to intradermal injection of lepromin (Mitsuda reaction) is linked to the human NRAMP1 gene in Vietnamese leprosy sibships. J. Infect. Dis. 181:302– 8 Alcais A, Remus N, El Baghdadi J, Abel L, Casanova JL. 2002. Genetic susceptibility to tuberculosis: from monogenic to polygenic inheritance. Sepsis. In press Griffin MD, Lutz W, Phan VA, Bachman LA, McKean DJ, Kumar R. 2001. Dendritic cell modulation by 1alpha,25 dihydroxyvitamin D3 and its analogs: a vitamin D receptor-dependent pathway that promotes a persistent state of immaturity in vitro and in vivo. Proc. Natl. Acad. Sci. USA 98:6800–5 Bellamy R, Ruwende C, Corrah T, McAdam KP, Thursz M, Whittle HC, Hill AV. 1999. Tuberculosis and chronic hepatitis B virus infection in Africans

188.

189.

190.

191.

192.

193.

194.

195.

619

and variation in the vitamin D receptor gene. J. Infect. Dis. 179:721–24 Wilkinson RJ, Llewelyn M, Toossi Z, Patel P, Pasvol G, Lalvani A, Wright D, Latif M, Davidson RN. 2000. Influence of vitamin D deficiency and vitamin D receptor polymorphisms on tuberculosis among Gujarati Asians in west London: a case-control study. Lancet 355:618–21 Stene LC. 2000. Vitamin D deficiency and tuberculosis. Lancet 356:73–74; Discussion: 4–5 Roy S, Frodsham A, Saha B, Hazra SK, Mascie-Taylor CG, Hill AV. 1999. Association of vitamin D receptor genotype with leprosy type. J. Infect. Dis. 179:187–91 Neth O, Jack DL, Dodds AW, Holzel H, Klein NJ, Turner MW. 2000. Mannosebinding lectin binds to a range of clinically relevant microorganisms and promotes complement deposition. Infect. Immun. 68:688–93 Selvaraj P, Narayanan PR, Reetha AM. 1999. Association of functional mutant homozygotes of the mannose binding protein gene with susceptibility to pulmonary tuberculosis in India. Tuberc. Lung Dis. 79:221–27 Bellamy R, Ruwende C, McAdam KP, Thursz M, Sumiya M, Summerfield J, Gilbert SC, Corrah T, Kwiatkowski D, Whittle HC, Hill AV. 1998. Mannose binding protein deficiency is not associated with malaria, hepatitis B carriage nor tuberculosis in Africans. Q. J. Med. 91:13–18 Hoal-Van Helden EG, Epstein J, Victor TC, Hon D, Lewis LA, Beyers N, Zurakowski D, Ezekowitz AB, Van Helden PD. 1999. Mannose-binding protein B allele confers protection against tuberculous meningitis. Pediatr. Res. 45:459–64 Garred P, Harboe M, Oettinger T, Koch C, Svejgaard A. 1994. Dual role of mannan-binding protein in infections: another case of heterosis? Eur. J. Immunogenet. 21:125–31

2 Feb 2002

11:3

Annu. Rev. Immunol. 2002.20:581-620. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

620

AR

AR152-19.tex

CASANOVA

¥

AR152-19.SGM

LaTeX2e(2001/05/10)

P1: GJC

ABEL

196. Juffermans NP, Florquin S, Camoglio L, Verbon A, Kolk AH, Speelman P, van Deventer SJ, van Der Poll T. 2000. Interleukin-1 signaling is essential for host defense during murine pulmonary tuberculosis. J. Infect. Dis. 182:902– 8 197. Wilkinson RJ, Patel P, Llewelyn M, Hirsch CS, Pasvol G, Snounou G, Davidson RN, Toossi Z. 1999. Influence of polymorphism in the genes for the interleukin (IL) -1 receptor antagonist and IL-1beta on tuberculosis. J. Exp. Med. 189:1863–74 198. Bellamy R, Beyers N, McAdam KP, Ruwende C, Gie R, Samaai P, Bester D, Meyer M, Corrah T, Collin M, Camidge

DR, Wilkinson D, Hoal-Van Helden E, Whittle HC, Amos W, van Helden P, Hill AV. 2000. Genetic susceptibility to tuberculosis in Africans: a genomewide scan. Proc. Natl. Acad. Sci. USA 97:8005–9 199. Siddiqui MR, Meisner S, Tosh K, Balakrishnan K, Ghei S, Fisher SE, Golding M, Shanker Narayan NP, Sitaraman T, Sengupta U, Pitchappan R, Hill AV. 2001. A major susceptibility locus for leprosy in India maps to chromosome 10p13. Nat. Genet. 27:439–41 200. Motulsky AG. 1989. Metabolic polymorphisms and the role of infectious diseases in human evolution. 1960. Hum. Biol. 61:835–69; Discussion: 70–77

P1: FRK

February 1, 2002

12:47

Annual Reviews

AR152-FM

Annual Review of Immunology Volume 20, 2002

Annu. Rev. Immunol. 2002.20:581-620. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

CONTENTS FRONTISPIECE, Charles A. Janeway, Jr. A TRIP THROUGH MY LIFE WITH AN IMMUNOLOGICAL THEME, Charles A. Janeway, Jr.

xiv 1

THE B7 FAMILY OF LIGANDS AND ITS RECEPTORS: NEW PATHWAYS FOR COSTIMULATION AND INHIBITION OF IMMUNE RESPONSES, Beatriz M. Carreno and Mary Collins

29

MAP KINASES IN THE IMMUNE RESPONSE, Chen Dong, Roger J. Davis, and Richard A. Flavell

PROSPECTS FOR VACCINE PROTECTION AGAINST HIV-1 INFECTION AND AIDS, Norman L. Letvin, Dan H. Barouch, and David C. Montefiori T CELL RESPONSE IN EXPERIMENTAL AUTOIMMUNE ENCEPHALOMYELITIS (EAE): ROLE OF SELF AND CROSS-REACTIVE ANTIGENS IN SHAPING, TUNING, AND REGULATING THE AUTOPATHOGENIC T CELL REPERTOIRE, Vijay K. Kuchroo, Ana C. Anderson, Hanspeter Waldner, Markus Munder, Estelle Bettelli, and Lindsay B. Nicholson

55 73

101

NEUROENDOCRINE REGULATION OF IMMUNITY, Jeanette I. Webster, Leonardo Tonelli, and Esther M. Sternberg

125

MOLECULAR MECHANISM OF CLASS SWITCH RECOMBINATION: LINKAGE WITH SOMATIC HYPERMUTATION, Tasuku Honjo, Kazuo Kinoshita, and Masamichi Muramatsu

165

INNATE IMMUNE RECOGNITION, Charles A. Janeway, Jr. and Ruslan Medzhitov

KIR: DIVERSE, RAPIDLY EVOLVING RECEPTORS OF INNATE AND ADAPTIVE IMMUNITY, Carlos Vilches and Peter Parham ORIGINS AND FUNCTIONS OF B-1 CELLS WITH NOTES ON THE ROLE OF CD5, Robert Berland and Henry H. Wortis E PROTEIN FUNCTION IN LYMPHOCYTE DEVELOPMENT, Melanie W. Quong, William J. Romanow, and Cornelis Murre

197 217 253 301

LYMPHOCYTE-MEDIATED CYTOTOXICITY, John H. Russell and Timothy J. Ley

SIGNAL TRANSDUCTION MEDIATED BY THE T CELL ANTIGEN x

323

P1: FRK

Annu. Rev. Immunol. 2002.20:581-620. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

February 1, 2002

12:47

Annual Reviews

AR152-FM

CONTENTS

xi

RECEPTOR: THE ROLE OF ADAPTER PROTEINS, Lawrence E. Samelson INTERACTION OF HEAT SHOCK PROTEINS WITH PEPTIDES AND ANTIGEN PRESENTING CELLS: CHAPERONING OF THE INNATE AND ADAPTIVE IMMUNE RESPONSES, Pramod Srivastava CHROMATIN STRUCTURE AND GENE REGULATION IN THE IMMUNE SYSTEM, Stephen T. Smale and Amanda G. Fisher PRODUCING NATURE’S GENE-CHIPS: THE GENERATION OF PEPTIDES FOR DISPLAY BY MHC CLASS I MOLECULES, Nilabh Shastri, Susan

371

Schwab, and Thomas Serwold

395 427

463

THE IMMUNOLOGY OF MUCOSAL MODELS OF INFLAMMATION, Warren Strober, Ivan J. Fuss, and Richard S. Blumberg T CELL MEMORY, Jonathan Sprent and Charles D. Surh

GENETIC DISSECTION OF IMMUNITY TO MYCOBACTERIA: THE HUMAN MODEL, Jean-Laurent Casanova and Laurent Abel ANTIGEN PRESENTATION AND T CELL STIMULATION BY DENDRITIC CELLS, Pierre Guermonprez, Jenny Valladeau, Laurence Zitvogel, Clotilde Th´ery, and Sebastian Amigorena

495 551 581

621

NEGATIVE REGULATION OF IMMUNORECEPTOR SIGNALING, Andr´e Veillette, Sylvain Latour, and Dominique Davidson

669

CPG MOTIFS IN BACTERIAL DNA AND THEIR IMMUNE EFFECTS, Arthur M. Krieg

PROTEIN KINASE Cθ

709 IN

T CELL ACTIVATION, Noah Isakov and Amnon

Altman

RANK-L AND RANK: T CELLS, BONE LOSS, AND MAMMALIAN EVOLUTION, Lars E. Theill, William J. Boyle, and Josef M. Penninger PHAGOCYTOSIS OF MICROBES: COMPLEXITY IN ACTION, David M. Underhill and Adrian Ozinsky

761 795 825

STRUCTURE AND FUNCTION OF NATURAL KILLER CELL RECEPTORS: MULTIPLE MOLECULAR SOLUTIONS TO SELF, NONSELF DISCRIMINATION, Kannan Natarajan, Nazzareno Dimasi, Jian Wang, Roy A. Mariuzza, and David H. Margulies

853

INDEXES Subject Index Cumulative Index of Contributing Authors, Volumes 1–20 Cumulative Index of Chapter Titles, Volumes 1–20

ERRATA An online log of corrections to Annual Review of Immunology chapters (if any, 1997 to the present) may be found at http://immunol.annualreviews.org/

887 915 925

8 Feb 2002

10:5

AR

AR152-20.tex

AR152-20.SGM

LaTeX2e(2001/05/10) P1: GJC 10.1146/annurev.immunol.20.100301.064828

Annu. Rev. Immunol. 2002. 20:621–67 DOI: 10.1146/annurev.immunol.20.100301.064828 c 2002 by Annual Reviews. All rights reserved Copyright °

Annu. Rev. Immunol. 2002.20:621-667. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

ANTIGEN PRESENTATION AND T CELL STIMULATION BY DENDRITIC CELLS Pierre Guermonprez1, Jenny Valladeau1, Laurence Zitvogel2, Clotilde Th´ery1 and Sebastian Amigorena1 1

Institut Curie, INSERM U520, 12 rue Lhomond, 75005 Paris, France, U520, Institut Curie; e-mail: [email protected], [email protected], [email protected], [email protected] 2 CNRS URA 1301, Laboratoire d’Immunologie Cellulaire, Institut Gustave Roussy; e-mail: [email protected]

Key Words T cell priming, immunotherapy, MHC, endocytosis, class presentation ■ Abstract Dendritic cells take up antigens in peripheral tissues, process them into proteolytic peptides, and load these peptides onto major histocompatibility complex (MHC) class I and II molecules. Dendritic cells then migrate to secondary lymphoid organs and become competent to present antigens to T lymphocytes, thus initiating antigen-specific immune responses, or immunological tolerance. Antigen presentation in dendritic cells is finely regulated: antigen uptake, intracellular transport and degradation, and the traffic of MHC molecules are different in dendritic cells as compared to other antigen-presenting cells. These specializations account for dendritic cells’ unique role in the initiation of immune responses and the induction of tolerance.

INTRODUCTION Dendritic cells represent a heterogenous cell population, residing in most peripheral tissues, particularly at sites of interface with the environment (skin and mucosae), where they represent 1%–2% of the total cell numbers (1, 2). In the absence of ongoing inflammatory and immune responses, dendritic cells constitutively patrol through the blood, peripheral tissues, lymph and secondary lymphoid organs. In peripheral tissues, dendritic cells take up self and nonself antigens. Internalized antigens are then processed into proteolytic peptides, and these peptides are loaded onto MHC class I and II molecules. This process of antigen uptake, degradation, and loading is called antigen presentation. Constitutively, however, peripheral dendritic cells present antigens quite inefficiently. A signal from pathogens, often referred to as a danger signal, induces dendritic cells to enter a developmental program, called maturation, which transforms dendritic cells into efficient antigen presenting cells (APCs) and T cell activators. Bacterial and viral products, as well as inflammatory cytokines and other self-molecules, induce dendritic cell maturation through direct interaction with 0732-0582/02/0407-0621$14.00

621

4 Feb 2002

11:53

Annu. Rev. Immunol. 2002.20:621-667. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

622

AR

AR152-20.tex

AR152-20.SGM

LaTeX2e(2001/05/10)

P1: GJC

GUERMONPREZ ET AL.

specific dendritic cell surface receptors. T lymphocytes, through CD40-dependent and -independent pathways, and endothelial cells contribute to the final maturation of dendritic cells through direct cell-to-cell contact and the secretion of cytokines (3). Soon after encountering a danger signal, the efficiency of antigen uptake, intracellular transport and degradation, and the intracellular traffic of MHC molecules are modified (4). Peptide loading as well as the half-life and delivery to the cell surface of MHC molecules is increased. The surface expression of T cell costimulatory molecules also rises. Thus, dendritic cells become the most potent APCs, and the only ones capable of activating naive T lymphocytes and of initiating adaptive immune responses. To interact with T cells, however, dendritic cells also need to migrate out of the tissues to reach secondary lymphoid organs. Concomitant with the modifications of their antigen presentation abilities, maturation also induces massive migration of dendritic cells out of peripheral tissues (2). Modifications in the expression of chemokine receptors and adhesion molecules, as well as profound changes of the cytoskeleton organization, contribute to the migration of dendritic cells, through the lymph, toward secondary lymphoid organs. By linking antigen uptake, peptide loading, and cell migration to the encounter of a danger signal (5), dendritic cells restrict antigen presentation to those antigens internalized during maturation, thus favoring the stimulation of T cells specific for potentially pathogenic antigens. It is important that, in addition to presenting antigens, dendritic cells also influence the outcome of immune responses (2). Different dendritic cell–subsets and dendritic cells at different maturation stages express distinct surface molecules and secrete different cytokines; thus they determine selectively the type of induced immune response. The extraordinary efficiency of dendritic cells for T cell stimulation prompted many groups around the world to undertake dendritic cell–based active immunotherapy protocols (6). As the fundamental mechanisms underlying antigen presentation in dendritic cells are being characterized, the information is used to optimize the preparation of dendritic cells, their sensitization with antigens, and the routes of injection. It has been fascinating in the last few years to unravel the cell biological basis of antigen presentation in dendritic cells and to witness almost instantaneously how the new information and concepts have been translated into the clinic. This review summarizes recent advances in the analysis of the cell biology of antigen uptake and presentation, as well as T cell stimulation by dendritic cells. We also discuss how these fundamental studies influenced the use of dendritic cells for active immunotherapy.

ANTIGEN AND PATHOGEN RECOGNITION AND UPTAKE IN DENDRITIC CELLS Dendritic cells were long believed to display low endocytic and phagocytic activities. Because of their inability to take up antigens, and in spite of their high MHC class II expression, dendritic cells were not considered as APCs. This idea lasted

4 Feb 2002

11:53

AR

AR152-20.tex

AR152-20.SGM

LaTeX2e(2001/05/10)

P1: GJC

ANTIGEN PRESENTATION BY DENDRITIC CELLS

623

Annu. Rev. Immunol. 2002.20:621-667. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

until phagocytic Langherans cells (LCs) were shown to be precursors of some dendritic cells found in lymphoid organs, and bone marrow-derived dendritic cells at an early stage of their development were shown to internalize particulate antigens (7). Tissue dendritic cells capture pathogens, infected cells, dead cells, or their derived products to use for antigen presentation. Interestingly, pathogen recognition and uptake are in many cases accompanied by activation/maturation of dendritic cells. Conversely, dendritic cells are also exploited by many pathogens as a route of entry for infection.

Receptor-Mediated Endocytosis Receptor-mediated endocytosis allows the uptake of macromolecules through specialized regions of the plasma membrane, termed coated pits. This process is initiated by a signal in the cytoplasmic tail of the endocytic receptor, which is recognized by a family of adaptors responsible for the recruitment of clathrin lattices and for the formation of clathrin-coated endocytic vesicles (8). A large number of endocytic receptors are selectively expressed by subpopulations of immature dendritic cells. FC, COMPLEMENT, HEAT SHOCK PROTEINS, AND SCAVENGER RECEPTORS Mouse immature dendritic cells express receptors for the Fc portion of immunoglobulins (FcR): Fcγ RI (CD64), Fcγ RIII (CD16), and Fcγ RII (CD32) (9, 10). In human, monocyte-derived dendritic cells express mainly Fcγ RII (CD32) and FcαR (CD89) (11); LCs express Fcγ RI and FcεRI (CD23) (12); and blood dendritic cells express Fcγ RII and Fcγ RI (9). The neonatal MHC class I–like FcR for IgG was also found on human monocyte-derived dendritic cells (13); another receptor of the immunoglobulin superfamily named immunoglobulin-like transcript (ILT)-3 was found in immature dendritic cells, and it allows antigen presentation (14). Immature dendritic cells express complement receptor CR3 and CR4, but not CR1 and CR2 (15). Heat shock proteins (hsps) derived from tumor cells or infected cells stimulate antigen-specific T cell responses in vivo. Hsc70 and gp96 bind APCs (B cells, macrophages, and dendritic cells) and are internalized through specific membrane receptors (16, 17). Both binding to cells and presentation of hsp-associated peptides were saturable and inhibited by excess of unlabelled hsp. Recently, CD91, the α2macroglobulin receptor, also named LRP (low density lipoprotein–related protein), was identified as a gp96 and hsc70 receptor on mouse APCs (18). Scavenger receptors (SRs) are cell surface glycoproteins defined by their potential to bind chemically modified low-density lipoproteins. SRs play an important role in host defense because they are implicated in internalization of various bacteria. SRs are classified according to their structure, and dendritic cells express at least one SR: CD36 (class B-SR), involved in the uptake of apoptotic bodies (19). Other recently identified SRs have yet to be analyzed on dendritic cells, such as SRCL, an SR with a lectin domain, which binds Escherichia coli and Staphylococcus aureus (20).

4 Feb 2002

11:53

624

AR

AR152-20.tex

AR152-20.SGM

LaTeX2e(2001/05/10)

P1: GJC

GUERMONPREZ ET AL.

C-type lectins bind ligands in a Ca++-dependent manner. They share a carbohydrate recognition domain (CRD) of 115–130 amino acids, containing four cysteins forming two disulfide bonds. Dendritic cells express several transmembrane C-type lectins, including type I multi-CRD lectins like the macrophage-mannose receptor (MMR) (21) or DEC205 (22) and type II singleCRD lectins such as CD23, the low-affinity IgE receptor (12). The MMR (mouse and human) is expressed on alveolar and differentiated macrophages, but not on freshly isolated blood monocytes. The MMR binds several monosaccharides as well as a wide variety of pathogen antigens, including yeast antigens (23), lipoarabinomannan (24), and desialylated immunoglobulins (25). The MMR is expressed on monocyte-derived dendritic cells, where it mediates antigen capture, transport to endosomes and lysosomes, and efficient antigen presentation (21). The MMR is also expressed on blood dendritic cells and interstitial dendritic cells in the dermis, but not on LCs. Mouse DEC205 was first described on mature dendritic cells, LCs, and thymic epithelial cells (22). In contrast to the MMR, which recycles between the plasma membrane and early endosomes, DEC205 follows an unusual intracellular pathway after internalization, passing through late endocytic compartment (26), thanks to a specific molecular motif (EDE) in DEC205 cytoplasmic tail, which allows efficient antigen presentation. In humans, DEC205 (also known as gp200MR6) was only reported in blood dendritic cells, and its function is not yet documented (27). Dendritic cell–specific ICAM3-grabbing nonintegrin (DC-SIGN) and Langerin are two type II lectins with mannose specificity, expressed on interstitial dendritic cells and LCs, respectively and exclusively (28, 29). Unlike DC-SIGN, which is more implicated in adhesion processes (see Pathogens Use Dendritic Cell Endocytic Receptors), Langerin induces the formation of a unique endocytic compartment of LCs, Birbeck granules (28).

Annu. Rev. Immunol. 2002.20:621-667. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

C-TYPE LECTIN FAMILY

Phagocytosis and Macropinocytosis Particulate and soluble antigens are efficiently internalized by phagocytosis and macropinocytosis, respectively. Both processes are actin dependent, require membrane ruffling, and result in the formation of large intracellular vacuoles. Phagocytosis is generally receptor mediated, whereas macropinocytosis is a cytoskeletondependent type of fluid-phase endocytosis. In macrophages and epithelial cells, macropinocytosis is transiently induced by growth factors or phorbol ester. In immature dendritic cells, in contrast, macropinocytosis is constitutive (21). Macropinocytosis represents a critical antigen uptake pathway allowing dendritic cells to rapidly and nonspecifically sample large amounts of surrounding fluid. Phagocytosis, in contrast, is initiated by the engagement of specific receptors, triggering a cascade of signal transduction, which is required for actin polymerization and effective engulfment. In general, the same receptors mediate both phagocytosis and clathrin-dependent endocytosis.

4 Feb 2002

11:53

AR

AR152-20.tex

AR152-20.SGM

LaTeX2e(2001/05/10)

P1: GJC

Annu. Rev. Immunol. 2002.20:621-667. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

ANTIGEN PRESENTATION BY DENDRITIC CELLS

625

PHAGOCYTOSIS OF PATHOGENS Immature dendritic cells were reported to phagocytose almost any bacteria: cocci GRAM+ (Streptococcus aureus and Streptococcus gordonii), bacilli GRAM+ (Listeria monocytogenes) and GRAM− (Salmonella typhimurium and Escherichia coli), and mycobacteria (Mycobacterium tuberculosis, BCG) (3). In gut epithelial monolayers, immature dendritic cells open tight junctions, project their dendrites to the apical side of the epithelia, and sample bacteria therein (30). Immature dendritic cells phagocytose Saccharomyces cerevisae and Candida albicans yeast cells and hyphae through two different pathways (coiling for yeast and zipper for hyphae) (31). Finally, murine spleen dendritic cells internalize parasites like Leishmania (L.) major, L. donovani, and L. mexicana (32, 33). PHAGOCYTOSIS OF APOPTOTIC AND NECROTIC BODIES Immature dendritic cells also internalize apoptotic and necrotic bodies. In vitro, apoptosis is induced by virus infection (influenza, vaccinia, measles, EBV, HIV1), X-ray-, γ - or UV-irradiation, serum starvation, Fas- or TRAIL-ligation, or by treatment with brefeldin A. In parallel, necrosis is generally obtained by freeze-and-thaw cycles or heating the cells at 50◦ C. Human monocyte–derived dendritic cells internalize apoptotic and necrotic bodies derived from T and B cell lines (34), virus-infected apoptotic monocytes (35), or tumor cell lines, including melanoma cells (36–38), squamous cell carcinoma (39), or kidney adenocarcinoma (40). Mouse bone marrow–derived dendritic cells engulf apoptotic bodies derived from fibroblasts (41), allogenic B cells (42), or macrophages infected with Salmonella typhimurium (43). Phagocytosis of apoptotic epithelial cells by murine LCs was also observed using electron microscopy on vaginal epithelium sections (44). Finally, immunohistology in rat lymph node, Peyer’s patches, and lamina propria also revealed cytokeratin-positive inclusions in dendritic cells, suggesting transport of apoptotic bodies by immature dendritic cells to T cell areas of lymph nodes (45). Molecules (solubles or receptors) mediating the engulfment of dying cells by macrophages have been extensively analyzed. Phagocytes use complement receptors, CD14, integrins (including αvβ3 and αvβ5, which can act in cooperation with CD36 and thrombospondin-1), SR-family members (like CD36, lox1, or class A-SR) (46), or the two recently described receptors PSR (47) and Mer (48). In dendritic cells, phagocytosis of apoptotic corpses occurs through a complex including CD36, αvβ5 or αvβ3, and signaling through the CrkII-Dock180-rac1 molecular complex (49–51).

Regulation of Endocytosis in Dendritic Cells Efficient antigen internalization is a specific attribute of immature dendritic cells. During maturation, dendritic cells downregulate their endocytic capacity, thus limiting the range of antigens that they will be able to present after leaving peripheral tissues. Downmodulation of internalization occurs through two independent mechanisms: a decrease in cell surface expression of most antigen receptors

4 Feb 2002

11:53

626

AR

AR152-20.tex

AR152-20.SGM

LaTeX2e(2001/05/10)

P1: GJC

GUERMONPREZ ET AL.

Annu. Rev. Immunol. 2002.20:621-667. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

(e.g., MMR/FcR) and the downmodulation of both macropinocytosis and phagocytosis (21, 52). Recently, two independent groups have analyzed the molecular mechanism of this process. They both found that inactivation of the small GTPases cdc42 and rac1 blocks macropinocytosis and phagocytosis in immature dendritic cells. In contrast to the findings of Garrett et al., West et al. could neither detect any changes in the active forms of cdc42 during maturation nor restore macropinocytosis in mature dendritic cells by microinjecting constitutively active cdc42. These results suggest that dendritic cells may downmodulate internalization through different mechanisms (53, 54).

Pathogens Use Dendritic Cell Endocytic Receptors It is interesting that pathogens learned how to use, for their own profit, unique dendritic cell characteristics—most notably their migratory capacity. Thus, immature dendritic cells and Langherhans cells play a critical role in the dissemination of viruses from mucosae to draining lymph nodes. In this context, different dendritic cell surface molecules serve as virus, parasite, or yeast receptors. Measles virus interacts with dendritic cells via FcRs or CD46 (55), when other viruses, like SV40, enter cat dendritic cells through caveolae (i.e., specialized cholesterol-rich membrane domain) (56). Parasites like Leishmania major infect LCs via CR3 (57), and Histoplasma capsulatum yeasts are phagocytosed via very late antigen-5 (58). The most extensively studied example of pathogen internalization in dendritic cells is HIV. The question of the susceptibility of dendritic cells to infection by HIV has been controversial, and dendritic cells now appear as Trojan Horses, transporting viruses to lymph nodes for T cell infection and virus dissemination (59). Interestingly, the HIV receptor (CD4) and coreceptor (CCR5 and CXCR4 chemokine receptors) are differentially expressed on dendritic cell subpopulations (60). CCR5 (a receptor for M tropic strains) is expressed on fresh LCs, whereas CXCR4 (a receptor for T tropic viruses) can be found on mature dendritic cells (61). In addition, interstitial dendritic cells, but not mucosal dendritic cells, express DC-SIGN, a surface dendritic cell molecule involved in HIV binding and trans-infection of T cells (62). The normal function of DC-SIGN is to help dendritic cell–T cell interactions and dendritic cell emigration out of blood or lymph vessels through its interaction with ICAM-3 and ICAM-2, respectively (29, 63).

INDUCTION OF DENDRITIC CELL MATURATION AND SIGNAL TRANSDUCTION The organism bears two arms to fight pathogens: innate and adaptive immunity. Innate immunity specifically recognizes conserved molecules of infectious antigens, absent from mammalian organisms (called pathogen-associated molecular patterns), through a family of specific receptors (pattern recognition receptors) (64). Pathogen recognition by the immune system has two major effects. First, it

4 Feb 2002

11:53

AR

AR152-20.tex

AR152-20.SGM

LaTeX2e(2001/05/10)

P1: GJC

Annu. Rev. Immunol. 2002.20:621-667. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

ANTIGEN PRESENTATION BY DENDRITIC CELLS

627

triggers effector cells of inflammation, including macrophages and polymorphonuclear neutrophils, which represent an immediate defense at the sites of pathogen entry. Second, the innate immune system induces adaptive immunity. Dendritic cells play a pivotal role as sensors of infection for the initiation of adaptive immune responses (65). Dendritic cells respond to two types of signals: direct recognition of pathogens (through specific pattern-recognition receptors) and indirect sensing of infection (through inflammatory cytokines, internal cellular compounds, and ongoing specific immune responses) (5). In response to these signals, dendritic cells are activated to enter an integrated developmental program called maturation, which transforms dendritic cells into efficient T cell stimulators. In this chapter, we briefly describe the receptors involved in the induction of dendritic cell maturation and the scarce information available concerning the signal transduction pathways triggered by these receptors. The cell biological modifications resulting in dendritic cell migration toward secondary lymphoid organs is only summarized. The functional consequences of dendritic cell maturation in terms of antigen presentation and T cell stimulation is discussed in subsequent chapters.

Receptors for Dendritic Cell Activation Five types of surface receptors were reported to trigger dendritic cell maturation: (i) Toll-like receptors (TLR), (ii) cytokine receptors, (iii) TNF-receptor (TNF-R) family molecules, (iv) FcR, and (v) sensors for cell death. (i) Dendritic cells mature in response to various pathogenic compounds, including several bacterial wall compounds (such as lipopolysaccharide—LPS), unmethylated CpG motifs, and double-stranded RNA. Most of these molecules are recognized by a large family of surface receptors called TLR (64). In different cells of the immune system, different pathogen-associated pattern molecules are specifically recognized by one or by a combination of TLRs. For example, TLR4 determines responses to GRAM− bacteria through binding to LPS; TLR2 is involved in responses to different GRAM+ cell wall compounds (including bacterial peptidoglycans), to bacterial lipoproteins, and to Klebsiella pneumoniae OmpA protein; TLR5 recognizes flagellin from both GRAM+ and GRAM− bacteria; and TLR9 binds to unmethylated CpG motifs (66). Dendritic cells express a subset of these molecules, including TLR2, TLR3, and TLR4 (67, 68). TLR2 triggers functional dendritic cell maturation in response to bacterial peptidoglycans (69), to lipopeptides (70), to mycoplasma lipoproteins (71), and to OmpA (72). It is interesting that TLR2 may be recruited to yeast and GRAM+ bacteriacontaining phagosomes of macrophages, which suggests that this family of receptors may also influence endocytic and/or phagocytic functions (73). (ii) Dendritic cells sense danger and infections indirectly through inflammatory mediators such as TNF-α, IL-1β, PGE-2, whose secretion is triggered by pathogens (2, 74).

4 Feb 2002

11:53

Annu. Rev. Immunol. 2002.20:621-667. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

628

AR

AR152-20.tex

AR152-20.SGM

LaTeX2e(2001/05/10)

P1: GJC

GUERMONPREZ ET AL.

(iii) CD4+ T cells induce dendritic cell maturation both in vivo [for the initiation of certain cytotoxic T lymphocytes (CTLs) responses; see DC Maturation and T Cell Priming] and in vitro. Triggering of CD40 by CD40L on T cells induces dendritic cell maturation (75, 76), but CD40-independent mechanisms also exist. Indeed, triggering of Fas and OX40L on dendritic cells by FasL and OX40, respectively, on T cells may induce functional maturation (77, 78) (see Do T Cells Regulate Dendritic Cell Half-Life?). (iv) Dendritic cells may also be activated through receptors for immunoglobulins (see Receptor-Mediated Endocytosis). Indeed, engagement of most FcR, including Fcγ RI, Fcγ RIII, Fcγ RI, and Fcγ R, by immune complexes or specific antibodies, induces dendritic cell maturation (11, 12, 79). This process requires the FcR-associated γ -chain that contains an immunoreceptor tyrosine-containing activation motif (ITAM). Tyrosine phosphorylation of this motif may initiate dendritic cell maturation. (v) Cell death can be sensed as a danger signal by dendritic cells. Shi et al. showed that cell injury releases adjuvant compounds that enhance T cell responses (80). Two studies showed that necrotic, but not apoptotic, cell death induces mouse and human dendritic cell maturation in vitro (40, 41), although another study found that apoptotic bodies may induce dendritic cell maturation (81). One should be careful in interpreting these results, however, because LPS, mycoplasma contamination (82) or CD40L expression by apoptotic bodies (83) could be responsible for dendritic cell maturation. The nature of the dendritic cell activating compounds is still unclear. Nucleotides, such as ATP and UTP, may activate dendritic cells through purinergic receptors (5). Hsps including gp96, hsp90, and hsc70 are released by necrotic cells and may induce dendritic cell maturation (84, 85). Although CD91 has recently been identified as a gp96, hsc70, and hsp90 receptor, its role in dendritic cell activation has not yet been explored. Hsp60 and hsc70 activate macrophages through CD14 and/or TLR (86–89), but receptors engaged on dendritic cells are not characterized. It is unlikely that all these pathways for dendritic cell maturation are redundant. Rather, they probably result in functionally distinct dendritic cell populations. In addition, different activation signals may act synergistically or regulate each other. For example, human TGF-β-untreated dendritic cells are activated by LPS or TNFα in contrast to TGF-β-treated monocyte-derived dendritic cells, which require a CD40-dependent signal to acquire high T cell stimulating activity (90).

Signal Transduction and Dendritic Cell Maturation Most of the receptors triggering dendritic cell maturation are expressed in other cell types, where the signal transduction cascades have been analyzed in detail. Few studies, however, have specifically analyzed signal transduction in dendritic cells. In a very elegant pioneering study, Rescigno et al. showed that LPS treatment

4 Feb 2002

11:53

AR

AR152-20.tex

AR152-20.SGM

LaTeX2e(2001/05/10)

P1: GJC

Annu. Rev. Immunol. 2002.20:621-667. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

ANTIGEN PRESENTATION BY DENDRITIC CELLS

629

of immature dendritic cells induces maturation through the activation of NF-κB and that it promotes dendritic cell survival by activating ERK (91). Ardeshna et al., more recently, confirmed that NF-κB is required for LPS-induced maturation of human monocyte-derived dendritic cells (92). They also found that survival of these dendritic cells depends on activation of the PI3K. CD40-mediated dendritic cell activation was also analyzed in terms of cell signaling (93, 94). CD40 signaling in dendritic cells is initiated within plasma membrane cholesterol-rich microdomains (rafts). TRAF2/3 and src-family protein kinases (such as Lyn) are recruited together with CD40 to membrane rafts. TRAF2/3 initiates the activation of p38MAPK, and lyn induces ERK activation (93). p38MAPK is required for CD40-induced IL-12 production in dendritic cells because dendritic cells from Mkk3 (a p38 activator)-deficient mice fail to produce IL-12 in response to CD40 stimulation (95). Calcium-ionophore treatment can also induce phenotypical and functional maturation of monocytic cells (96). Although LPS, TNF-α and calcium ionophore signaling pathways all lead to NF-κB activation, only the calcium ionophore involves calmodulin/calcineurin activation (96). Recent results point at MyD88 as an essential component of TLR signaling in dendritic cells (97). Not all TLR, however, activate a MyD88-dependent signaling pathway. In the absence of MyD88, TLR2- and TLR9-induced dendritic cell maturation is completely abolished, whereas TLR4, after LPS binding, can still induce full dendritic cell maturation (98, 99). A MyD88-independent pathway involving MAPK and NF-κB pathways therefore exists downstream of TLR4 (98). Finally, FcεRI triggering in LCs induces phosphorylation of the tyrosine kinase syk and Ca++ mobilization (12). We found recently that Fcγ R engagement in mouse dendritic cells induces syk and ERK phosphorylation and that syk is indispensable for immune complex–induced dendritic cell maturation (C. Sedlik, D. Orbach, E. Schweighoffer, F. Colucci, S. Verbeek, P RicciardiCastagnoli, V.L.J. Tybulewicz, J. Di Santo, S. Amigorena, in preparation). The process of dendritic cell maturation is intimately linked to their migration out of peripheral tissues toward lymph nodes. Dendritic cell migration is due to coordinated changes in dendritic cell adhesion properties (changes in cytoskeleton organization and integrin expression were reported) and in chemokine receptor expression (100). Importantly, trans-endothelial migration itself participates in dendritic cell differentiation and maturation (101). Although many studies have analyzed dendritic cell maturation in vitro and in vivo, a coherent picture reconciling all these studies with the physiological process of dendritic cell maturation has yet to emerge. The precise coordination between the phenotypical, morphological, and functional modifications induced during maturation and the migration of dendritic cells is unclear. It was initially proposed that only fully mature dendritic cells migrate efficiently. However, it is now clear that in many cases final maturation requires dendritic cell interaction with T cells. It is unlikely that dendritic cells encounter specific T cells in the periphery. Thus, it is most likely that dendritic cells leave peripheral tissues in an intermediate stage of maturation and

4 Feb 2002

11:53

630

AR

AR152-20.tex

AR152-20.SGM

LaTeX2e(2001/05/10)

P1: GJC

GUERMONPREZ ET AL.

become fully mature in the draining lymph nodes, under the control of specific T cells.

Annu. Rev. Immunol. 2002.20:621-667. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

ANTIGEN PRESENTATION AND CROSS-PRESENTATION CD8+ and CD4+ T cells express clonally distributed receptors that recognize fragments of antigens (peptides) associated with MHC class I and II molecules, respectively. Antigen degradation and peptide loading onto MHC molecules occur intracellularly in APCs. In the last 20 years, the intracellular pathways for peptide loading on MHC class I and II molecules were analyzed in detail. A strict compartmentalization of MHC class I and II biogenesis results in the loading of exogenous and endogenous antigens on MHC class II molecules in the endocytic pathway, and the selective loading of endogenous, but not exogenous, antigens on MHC class I molecules in the endoplasmic reticulum (ER). This model accounts, at the effector level, for the selective killing by MHC class I–restricted CD8+ CTLs of virus-infected cells (expressing endogenous viral antigens), but not of neighboring cells that have internalized inactive virus or apoptotic infected cells. This model, however, is also in conflict with experiments from M. Bevan, who demonstrated, over 20 years ago, that priming of CTL immune responses in vivo can occur after presentation of exogenous antigens by MHC class I molecules (a phenomenon that he called cross priming) (102). Recent studies in dendritic cells (and other APCs) reconcile these two series of studies by showing that, in addition to MHC class II, internalized antigens may also be presented by MHC class I molecules (a phenomenon referred to as cross presentation).

MHC Class I–Restricted Antigen Presentation in Dendritic Cells PRESENTATION OF ENDOGENOUS ANTIGENS Most peptides to be loaded on MHC class I molecules are generated by proteasome degradation of newly synthesized ubiquitinated proteins. The resulting peptides are transferred to the ER by specialized peptide transporters, TAP, and loaded on new MHC class I molecules under the control of a loading complex composed of several ER resident chaperons (including tapasin, calnexin, calreticulin) (103). Once associated to peptides, MHC class I molecules are rapidly transferred through the Golgi apparatus to the plasma membrane. Like other cells, dendritic cells present self- or virus-derived endogenous antigens. Few studies have analyzed MHC class I biogenesis and endogenous peptide loading in dendritic cells. It was shown, however, that MHC class I synthesis and half-life are increased upon induction of dendritic cell maturation (104), although less strongly than those of MHC class II (105). It is interesting that, in contrast to MHC class II, MHC class I molecules are still efficiently synthesized and transported to the cell surface in mature dendritic cells (104, 106), again highlighting the functional differences in the regulation of antigen presentation to CD4+ and CD8+ T cells. In addition, dendritic cells constitutively express low levels of

4 Feb 2002

11:53

AR

AR152-20.tex

AR152-20.SGM

LaTeX2e(2001/05/10)

P1: GJC

ANTIGEN PRESENTATION BY DENDRITIC CELLS

631

Annu. Rev. Immunol. 2002.20:621-667. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

the immunoproteasome, which becomes the main type of proteasome in mature dendritic cells (107–109): LMP2, LMP7, and PA28 are all induced during dendritic cell maturation. This change in proteasome composition may upregulate the efficiency of presentation of some epitopes while decreasing presentation of others (109). Targeting of proteins for proteasomal degradation requires their ubiquitination. Interestingly, dendritic cells express a particular di-ubiquitin gene, encoded within the MHC locus, which may participate in antigen ubiquitination (110). WHICH INTERNALIZATION PATHWAYS LEAD TO CROSS PRESENTATION? Exogenous antigens internalized by various pathways, however, may also be presented by MHC class I molecules. Macropinocytosis allows receptor-independent cross presentation of soluble antigens by dendritic cells (111). Physiologically, phagocytosis is probably a major route for antigen uptake and cross presentation. Pioneering work in macrophages, and more recently in dendritic cells, showed that linking antigens to latex beads, thus forcing internalization by phagocytosis, strongly increased the efficiency of cross presentation (112, 113). Whether efficient cross presentation in this case is exclusively due to phagocytosis remains unclear since (a) a soluble antigen co-internalized with the beads was also cross presented (114) and (b) the same antigen coupled to red blood cells, which are also phagocytosed, did not induce cross presentation (115). These results suggest that the latex beads could somehow disrupt endocytic compartments, thus favoring cytosol delivery and cross presentation. Phagocytosis of bacteria results in efficient cross presentation in both macrophages and dendritic cells (104, 116). Interestingly, bystander dendritic cells and macrophages phagocytosed Salmonella-infected cells efficiently, but only dendritic cells cross presented bacteria-derived antigens (43). Phagocytosis of apoptotic cells also results in efficient cross presentation of viral (34, 35, 117) and tumor antigens (37, 38, 118). Whether apoptosis per se is required for the cross presentation process is a matter of debate. Albert et al. showed that death by necrosis or inhibitors of apoptosis resulted in inefficient cross presentation (35). Cross presentation after uptake of necrotic cells or of cellular lysates, however, may also occur (34, 119). We found that exosomes, small membrane vesicles secreted by tumor cells, contain tumor antigens and that their uptake by dendritic cells results in cross presentation (120). In addition, FcR-mediated uptake of immune complexes (79), opsonized liposomes (121), or opsonized dead cells (122) promotes efficient cross presentation. In human monocytes, antigen targeting to Fcγ RI also resulted in cross presentation (123). Cross presentation after FcR-mediated uptake is observed at antigen concentrations 3 to 5 logs below cross presentation after fluid phase uptake (79, 121). This cross presentation pathway is strictly dependent on immune complex receptors since dendritic cells from receptor-deficient mice failed to cross present immune complexes efficiently (79). Peptides bound to different hsp, including gp96, hsp90, hsp60, and hsc70, induce efficient cross presentation both in vitro and in vivo. At steady state, hsps are

4 Feb 2002

11:53

Annu. Rev. Immunol. 2002.20:621-667. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

632

AR

AR152-20.tex

AR152-20.SGM

LaTeX2e(2001/05/10)

P1: GJC

GUERMONPREZ ET AL.

loaded with endogenous peptides. Loading of gp96 with peptides, for example, occurs in the ER and is TAP dependent (124). Hsps are released upon necrotic cell death (84) and may be internalized by APCs through specific receptors, such as CD91 (16, 18). Hsp-associated peptides are then somehow transferred to MHC class I and class II molecules for antigen presentation (125, 126). Blocking hsp uptake through CD91 prevents cross presentation, suggesting that receptor-mediated uptake is required (125). Antigens associated to different bacterial products [including OmpA (72), Shiga toxin (127), and the CyaA toxin (128)] too, enter cells through specific receptors and induce efficient cross presentation. Whether receptors only accumulate antigens within dendritic cells or whether cross presentation requires targeting of antigens to specific endocytic compartments is unknown. The wide variety of receptors that seem to promote cross presentation could favor the first possibility. It is possible, however, that receptors selectively drive antigens to specific endocytic cross presentation compartments. HOW ARE INTERNALIZED ANTIGENS CROSS-PRESENTED? Two main intracellular pathways for cross presentation were reported, resulting in either endocytic or ER MHC class I peptide loading (129). Loading in endocytic compartments is in general insensitive to inhibitors of protein neosynthesis (brefeldineA or cycloheximide) and to inhibitors of the proteasome (lactacystine). In addition, loading is independent of the TAP transporters and sensitive to inhibitors of lysosomal function (ammonium chloride, chloroquine, bafilomycinA, etc.). Conversely, ER peptide loading is blocked by proteasome inhibitors and requires the expression of TAP. This pathway requires transport of internalized antigens from endocytic compartments to the cytosol. TAP-independent endocytic cross presentation requires the presence of MHC class I molecules in the endocytic pathway. Indeed, recycling MHC class I molecules are found in endosomes and lysosomes in B cells and dendritic cells (108, 130, 131). In dendritic cells, like class II molecules, MHC class I molecules are redistributed to the plasma membrane upon induction of maturation (108). TAPindependent endocytic cross presentation was reported for various antigen forms in B cells (130) and macrophages (116). In U937 cells, internalized Fcγ RI colocalizes with MHC class I molecules in rab4/5+, Lamp1− endocytic compartments, suggesting that peptide loading may occur therein (132). Finally, endocytic loading of MHC class I molecules in macrophages was directly evidenced using MHC class I/peptide-specific antibodies (16). Therefore, it is most likely that low endosomal pH allows peptides generated in endosomes to exchange with endogenous peptides bound to recycling MHC class I molecules. TAP-dependent cross presentation was reported in dendritic cells in many different experimental systems. Cross presentation of soluble OVA, latex beadbound OVA, and immunoglobulin-complexed OVA occurs through a proteasomesensitive, TAP-dependent pathway in mouse dendritic cells (79, 111, 133). Cross presentation of gp96-associated peptides is also TAP dependent (18). Hsc70-bound peptides may be presented in macrophages through TAP-dependent

4 Feb 2002

11:53

AR

AR152-20.tex

AR152-20.SGM

LaTeX2e(2001/05/10)

P1: GJC

Annu. Rev. Immunol. 2002.20:621-667. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

ANTIGEN PRESENTATION BY DENDRITIC CELLS

633

or -independent pathways according to the sequence context of the peptides (16). In this case, ER loading of the exogenous antigen was shown morphologically, using MHC class I/peptide-specific monoclonal antibodies (16). Finally, tapasinedeficient dendritic cells failed to cross present OVA to specific T cells, confirming that peptide loading after antigen internalization may occur in the ER (134). The TAP dependency of cross-presentation led K. Rock et al. to propose the existence of a membrane transport pathway linking the lumen of endocytic compartments and the cytosol (133). The existence of such a transport pathway was evidenced in macrophages (133, 135) and dendritic cells (111, 136). Indeed, dendritic cells bear a unique endosome-to-cytosol transport pathway that allows selective delivery of internalized antigens to the cytosol (136). Other cells like thymic epithelial (137) or liver sinus endothelial cells (138) also present, although inefficiently, soluble OVA by MHC class I in a TAP-dependent fashion. It is most likely that, depending on the antigen, the cell type, and the route of internalization, peptide loading may occur in endosomes, the ER, or both. It is interesting, nevertheless, that in vivo cross priming was mostly dependent on TAP (139, 140), suggesting that antigen delivery to the cytosol might be required for the initiation of CTL immune responses in vivo (see T Cell Stimulation by Dendritic Cells).

MHC Class II–Restricted Antigen Presentation in Dendritic Cells Peptide loading on MHC class II molecules occurs through a different pathway. Soon after synthesis in the ER, three α/β MHC class II dimers associate to a trimer of invariant (Ii) chains (141). These nonamers exit the ER and pass through the Golgi apparatus before being transported to the endocytic pathway, under the influence of transport signals present in the cytoplasmic region of the Ii chain. Once in endosomes and lysosomes, the nonameric complexes meet an acidic, protease-rich environment, where the Ii chain is degraded by several proteolytic enzymes of the cathepsin family (142). MHC class II dimers become competent to bind antigenic peptides under the control of two nonpolymorphic MHC class II molecules HLA-DM/H2-M and HLA-DO/H2-O (in human/mouse) (143). Once loaded with peptides, Ii chain-free MHC class II/peptide complexes reach the plasma membrane. Antigen degradation in the endocytic pathway and the generation of antigenic peptides require several proteases, including cathepsins and asparaginyl endopeptidases (144). Thus, MHC class II molecules bind mainly to peptides derived from antigens present in the endocytic pathway (internalized or membrane proteins), although many reports show endogenously synthesized antigens presented on MHC class II molecules. A specific feature of dendritic cells resides in the tight regulation of the cell surface display of MHC class II/peptide complexes during dendritic cells’ life cycle. Immature dendritic cells expose very few MHC class II/peptide complexes at their surface. Several intracellular mechanisms cooperate to achieve this phenotype.

4 Feb 2002

11:53

Annu. Rev. Immunol. 2002.20:621-667. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

634

AR

AR152-20.tex

AR152-20.SGM

LaTeX2e(2001/05/10)

P1: GJC

GUERMONPREZ ET AL.

First, antigen degradation is inefficient in immature dendritic cells, and internalized antigens can remain intact in lysosomal compartments for several days (145); the availability of antigenic peptides to load on MHC class II molecules is therefore limited. Poor antigen degradation in immature dendritic cells is due to a low efficiency of several proteases, including cathepsin B, one of the major proteases involved in antigen degradation in dendritic cells (146). A splice variant of Ii, Iip41, known to inhibit certain cysteine proteases (147) and detected in immature dendritic cell lysosomes in association with H2-M (148), could contribute to the poor protease activity of immature dendritic cells. Second, even if a few antigenic peptides are formed in endosomes and lysosomes, very few MHC class II molecules are available to bind them in immature dendritic cells. In some murine bone marrow–derived dendritic cells, MHC class II haplotypes with a strong affinity for Ii remain associated in lysosomes with a partially degraded form of Ii, termed Iip10, that blocks access to the peptide binding groove (149, 150). Low protease activity of cathepsin S, the main protease for Iip10 degradation in dendritic cells (151–153), is involved here, and a role of the cathepsin S inhibitor cystatin C, specifically localized in lysosomal compartments together with cathepsin S, Ii, and MHC class II in immature dendritic cells, has been proposed (150). This type of regulation, however, is not general to all dendritic cells because it has not been observed in immature murine dendritic cells expressing other MHC haplotypes with lower affinity for Ii (151, 154), nor in murine dendritic cells obtained from different sources or cultured in different conditions (155), nor in human dendritic cells (105, 146). In these cases, Iip10 is degraded in lysosomes, and MHC class II dimers then reach the cell surface in association with either antigenic peptides, or the CLIP peptide derived from Ii (105, 151, 155), or without any peptide (156). Once at the cell surface, MHC class II molecules are rapidly internalized and can associate with new peptides in recycling endosomes before going back to the cell surface (156, 157), or they are directed to lysosomes, where they will finally be degraded (105, 155). This pathway results in a short-term presentation of MHC/peptide complexes at the immature dendritic cell surface, which is probably not sustained enough to stimulate T cells efficiently. Maturation signals induce a coordinate modification of all these aspects of MHC/peptide complex formation and transport. MHC class II synthesis is transiently upregulated (104, 105), and protease activity of the cathepsins involved in antigen and in Ii degradation quickly increases, maybe due to their relocalization to more acidic compartments (146). As a result, the numbers of available peptides and of MHC dimers free to form complexes increase concomitantly. MHC/peptide complexes are then rapidly formed (145) and transported to endosomal vesicles (149), where they colocalize with costimulatory (B7-2) and MHC class I molecules before being delivered to the cell surface as clusters of molecules involved in T cell stimulation (158). During maturation, the endocytosis activity also decreases, and transport of internalized MHC class II molecules to lysosomes for degradation is greatly reduced, resulting in stabilization of MHC class II/peptide complexes

4 Feb 2002

11:53

AR

AR152-20.tex

AR152-20.SGM

LaTeX2e(2001/05/10)

P1: GJC

Annu. Rev. Immunol. 2002.20:621-667. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

ANTIGEN PRESENTATION BY DENDRITIC CELLS

635

at the cell surface (105, 155). Later on, MHC class II synthesis is downregulated, and association of peptides with newly synthesized MHC molecules becomes very inefficient, focusing the range of peptides displayed at the mature dendritic cell surface to those arising from antigens encountered before or during the induction of maturation. Interestingly, however, fully mature dendritic cells can still form new MHC/peptide complexes using MHC molecules recycling from the cell surface (157). Anti-inflammatory cytokines can interfere with the regulation of MHC class II processing pathways in dendritic cells. M-CSF induces a rapid and sustained upregulation of MHC class II synthesis and transport to the cell surface in human immature dendritic cells, but without stabilizing them at the cell surface; the resulting dendritic cells are therefore very inefficient T cell stimulators (159). IL-10 inhibits the rise in protease activity induced by maturation signals by increasing the endosomal pH, and therefore it induces a long-term inhibition of antigen processing and presentation by mature dendritic cells (146). The cytokine environment of dendritic cells, when they receive a danger signal, is therefore a crucial factor in determining the orientation of the induced immune response.

CD1-Restricted Antigen Presentation In addition to MHC class I and II, dendritic cells express a third class of MHC molecules involved in antigen presentation to T cells: the CD1 proteins (160). Five CD1 genes encoding for related nonpolymorphic proteins (CD1a, b, c, d, e) have been identified in human, and only two in mouse, both homologues of CD1d. On the bases of sequence homologies, CD1a, b, c, and e were classified as group 1, and CD1d as group 2 CD1 molecules. Whereas CD1a, b, and c are mainly expressed by dendritic cells and cortical thymocytes, CD1d is expressed in various hematopoietic cells (dendritic cells, thymocytes, circulating B cells, T cells, and monocytes) and in intestinal epithelium. Like MHC class I, CD1 molecules are noncovalently associated with β2-microglobulin. Group 1 CD1 molecules, complexed with glycolipidic antigens of either endogenous (self-lipids, e.g., GM1) or exogenous (e.g., mycobacteria-derived lipids) origin, stimulate T cells of various phenotypes: CD8+ cytotoxic cells, CD4−CD8− T cells, and γ /δ T cells (161). CD1/antigen association takes place mostly in compartments of the endocytic pathway (162), where CD1 molecules are transported after internalization from the plasma membrane. Each CD1 molecule surveys specific endosomal compartments: early and recycling endosomes for CD1a (163, 164), lysosomes for CD1b, and probably all compartments for CD1c (165). CD1a, b, and c also meet phagocytosed mycobacteria in phagosomes of different maturation stages (166). Finally, antigen association with CD1b and CD1c can also take place at the dendritic cell surface (167, 168). In human and mouse, the group 2 CD1d molecule complexed with a synthetic lipid, α-galactosyl-ceramide (α-GalCer), activates NKT cells, a subset of T cells of either CD4+ or CD4−CD8− phenotype that express an invariant TCR and the

4 Feb 2002

11:53

Annu. Rev. Immunol. 2002.20:621-667. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

636

AR

AR152-20.tex

AR152-20.SGM

LaTeX2e(2001/05/10)

P1: GJC

GUERMONPREZ ET AL.

NK1.1 receptor (169). The current idea is that α-GalCer mimics an antigenic lipid of cellular origin (170), which associates with CD1d in the endocytic pathway (171). Other T cells, negative for the invariant TCR and NK1.1, recognize CD1d molecules that have formed a complex with endogenous lipids inside the ER (171). Very hydrophobic peptides may also form complexes with CD1d (172). Scarce data on the specific function of CD1d expression on dendritic cells are available. In human, NKT cells stimulated via CD1d on myeloid dendritic cells induce dendritic cell lysis (173), thereby probably modulating the normal immune response. In mice, CD1d-mediated interaction between NKT cells and αGalCer-pulsed dendritic cells induces IL-12 production by dendritic cells (174) and subsequent activation of both CD4+ and CD8+ T cells (175). We have recently observed that stressed murine dendritic cells, because they upregulate B7, can activate NKT cells by presenting self-antigens in a CD1d context, whereas in resting dendritic cells, inhibitory signals delivered upon MHC class I recognition counteract the CD1d-dependent NKT cell stimulation (176). The fifth CD1 gene encodes for multiple isoforms of the CD1e protein, none of which are expressed at the cell surface (177). One isoform accumulates in the Golgi apparatus and is relocalized to late endosomes upon dendritic cell maturation (177). The function of CD1e remains to be determined.

T CELL STIMULATION BY DENDRITIC CELLS Priming and Tolerization Several lines of evidence define dendritic cells as the principal professional APCs for T cell priming. On one hand, dendritic cells were directly compared to other cells in in vitro assays for the priming of alloreactive, na¨ıve TCR transgenic T cells or the expansion and activation of antigen-specific na¨ıve precursors from polyclonal populations (1). On the other hand, in vivo, injection of antigen-loaded dendritic cells induces potent CD4+ and CD8+ T cell primary responses (178, 179). The antigen presenting capacity of dendritic cells in situ was also assessed: Specific MHC/peptide complexes were probed (by T cells and, later on, by specific antibodies) at the surface of the ex vivo purified dendritic cells recovered from mice infected (180, 181) or immunized with soluble proteins in adjuvants (182– 185) and DNA (186, 187). Jenkins’s group directly visualized the interaction between antigen-specific transgenic T cells and antigen-loaded dendritic cells by immunofluorescence on lymph node sections (188). Dendritic cells were even shown to break T cell neonatal tolerance (189), peripheral tolerance against soluble antigens (190), transplantion antigens (191), peripheral tissue antigens (192, 193), and tumor antigens (194, 195). Importantly, dendritic cells are also involved in central and peripheral tolerance induction. The presence of dendritic cells in the thymus medulla suggested that they may participate in the establishment of central tolerance. Indeed, Brocker et al. showed that the exclusive expression of MHC class II in dendritic cells

4 Feb 2002

11:53

AR

AR152-20.tex

AR152-20.SGM

LaTeX2e(2001/05/10)

P1: GJC

ANTIGEN PRESENTATION BY DENDRITIC CELLS

637

Annu. Rev. Immunol. 2002.20:621-667. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

was sufficient to trigger Vβ-specific deletion of T cells by a retrovirally encoded superantigen (196). Using ex vivo detection of MHC/peptide complexes formed in vivo, several studies demonstrated that thymic (197) and splenic dendritic cells (184, 198, 199), beside B cells, are involved in antigen presentation after intravenous injection of tolerogenic high antigen doses. Mature dendritic cells from the T cell areas of lymph nodes present self-antigen in the periphery (183, 200). In addition, antibody-mediated targeting of antigen to some surface dendritic cell molecules induced specific T cell tolerance (201, 201a). Finally, dendritic cells could also participate in the induction of tolerance against liver allograft (202).

Cross Priming and Cross Tolerization Shortly after the discovery of MHC restriction, M. Bevan characterized the cross priming of CTLs against cell-associated exogenous antigens (see Antigen Presentation and Cross Presentation) (102). Examples of productive immunization against cellular antigens absent from APCs were obtained in the context of grafts, tumors, and viruses. In the case of tumors, Huang et al. showed that (a) the APC involved in this phenomenon derive from the bone marrow (203) and (b) the TAP transporters are required for priming, suggesting an access of the cell-associated antigen to the cytosol (139). TAP dependence of cross priming was recently extended to various viral infections (140), although some epitopes bypassed presentation by bone marrow cells or TAP requirement (204, 205). Some of the studied viruses (like influenza) may, however, infect hematopoietic cells, bringing into question the exclusive involvement of cross priming in the analyzed responses. Dendritic cells are also responsible for induction of cross tolerization (i.e., the induction of CTL tolerance against antigens that are not expressed in dendritic cells). Kurts et al. extensively described the existence of constitutive cross presentation between a model antigen expressed in the pancreas and bone marrow–derived APCs in the draining lymph nodes. This cross presentation results in transient proliferation and subsequent Fas-dependent deletion of autoreactive T cells (206). Expression of the relevant MHC class I allele selectively in dendritic cells was sufficient to support cross tolerization (207). The implication of tumor-infiltrating dendritic cells in cross presentation of tumor antigens has also been demonstrated (208). The mechanisms involved in cell-associated antigen transfer to dendritic cells are still unclear. It is most likely that antigen uptake by dendritic cells occurs in peripheral tissues before their migration to the draining lymph nodes. It is possible, although not demonstrated, that cross presentation requires phagocytosis of antigen-bearing cells since cross presentation and cross priming were more efficient using cell-associated antigens than using the corresponding soluble protein (for MHC class I cross presentation) (209) or peptide (for MHC class II presentation) (42). The induction of cross presentation in low OVA-expressing mice upon tissue destruction (by autoreactive T cells) supports this view (206). Alternative mechanisms like capture of hsp released from necrotic cells and/or capture of

4 Feb 2002

11:53

Annu. Rev. Immunol. 2002.20:621-667. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

638

AR

AR152-20.tex

AR152-20.SGM

LaTeX2e(2001/05/10)

P1: GJC

GUERMONPREZ ET AL.

membrane vesicles (exosomes) secreted by living cells may also promote cross presentation in vivo (see Antigen Presentation and Cross Presentation). Lymph node–resident dendritic cells might also cross present migrating dendritic cells. This is suggested by the efficient presentation of antigens present on subcutaneously injected dendritic cells by MHC class II from host-derived mature dendritic cells in the T cell areas of the draining lymph nodes (42). Surprisingly, however, in F1 mice immunized with parental MHC-bearing dendritic cells, the responses are mainly restricted by the MHC allele from the injected dendritic cells, indicating that this phenomenon does not account for efficient T cell priming in vivo (178).

Dendritic Cell Subsets and T Cell Priming The heterogeneity of dendritic cells in mouse and humans has been reviewed recently (210). In mouse, as originally stated by Shortman’s group, “lymphoid” CD11c+CD8α +DEC205+CD11b− and “myeloid” CD11c+CD8α −DEC205− CD11b+ dendritic cells can be distinguished (211). The lymphoid versus myeloid distinction was recently challenged by the demonstration that both common myeloid and lymphoid precursors can give rise to both subsets (the latter also gives rise to LCs) (211). Splenic CD8α + and CD8α − dendritic cells do not exhibit significant differences in their ability to stimulate CD8+ na¨ıve T cells in vitro and after in vivo intravenous injection (212). Recently, MHC/peptide complexes in splenic dendritic cells cross presenting soluble or cell-associated OVA were analyzed (213, 214). CD8α +CD11b− dendritic cells were more efficient for MHC class I cross presentation (213, 214) and CD8α −CD11b+ dendritic cells for MHC class II–restricted presentation (213). Because the two subsets of spleen dendritic cells do not differ strikingly in their uptake abilities, the inability of CD8α −CD11b+ dendritic cells to cross present antigen could be due to processing defects or, alternatively, to a difference in the accessibility of the two subsets to injected antigens. This apparent specialization of two dendritic cell subsets for MHC class I- and II-restricted antigen presentation is surprising because antigen presentation through both pathways by the same dendritic cells is required for effective CTL responses. In contradiction with these results, Reis e Sousa et al. showed that intravenous administration of lysosyme leads to MHC class II presentation mostly by CD8α +, but also by CD8α − splenic dendritic cells (185). Subcutaneous injection of antigen-loaded CD8α + dendritic cells primed Th1 responses, whereas CD8α − dendritic cells primed Th2 responses (215). These results are consistent with the responses obtained in mice where dendritic cell subsets were expanded by growth factor administration before immunization (216). These studies also challenge the view that CD8α + dendritic cells induce CD4+ T cell deletion (217). Intriguingly, subcutaneously injected CD8α + dendritic cells do not reach draining lymph nodes efficiently, although they efficiently prime CD4+ T cells by an unknown mechanism (218). Orientation of T cell responses by CD8α + and CD8α − dendritic cells is directly related to the cytokines produced by these dendritic cell subsets (219). The selective

4 Feb 2002

11:53

AR

AR152-20.tex

AR152-20.SGM

LaTeX2e(2001/05/10)

P1: GJC

Annu. Rev. Immunol. 2002.20:621-667. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

ANTIGEN PRESENTATION BY DENDRITIC CELLS

639

induction of Th1 responses by CD8α + dendritic cells is in direct causal relationship with their selective ability to produce IL-12 (215). Indeed, CD8α + dendritic cells secrete more bioactive IL-12 than their CD8α − counterparts; it should be noted that IL-4 upregulates IL-12 secretion in both subsets (219). The nature of the IFN-γ secreting subset remains controversial (219). In humans, besides the GM-CSF-dependent, monocyte-derived dendritic cells from myeloid origin, Grouard et al. characterized a CD4+CD11c− dendritic cell precursor in the T cell area of tonsils ultrastructurally as resembling plasmocytes (220). These plasmacytoid dendritic cells (pDC or DC2), unlike monocyte-derived dendritic cells (or DC1), are unresponsive to GM-CSF, but respond to IL-3 and express a specific combination of surface ILT (ILT1−ILT3+) (210). DC2 are also found in thymus and blood and seem to be of lymphoid origin (210). When cultured in IL-3, these dendritic cells activate Th2 cells, whereas monocyte-derived dendritic cells activate Th1 cells (221).

Dendritic Cell Maturation and T Cell Priming Maturation is associated with a rapid relocalization of antigen-bearing dendritic cells to the T cell zone of secondary lymphoid organs. Maturation also enhances surface levels of costimulatory and adhesion molecules such as B7.1 and B7.2, which bind both CD28 and CTLA-4 on T cells. These modifications increase the T cell priming ability of dendritic cells. If CD28 is constitutively expressed and facilitates the initial phases of T cell priming, CTLA-4 is inducible and is involved in the downregulation of T cell responses. Recently, it was shown that PD-1, an inhibitory receptor expressed on T cells, binds to PD-L1/B7-H1, a member of the B7 family expressed notably in human and murine mature dendritic cells (222). According to Dong et al. (223), B7H-1 costimulation leads to the secretion of IFN-γ and IL-10, whereas Freeman et al. (222) described a decrease in T cell proliferation and cytokine secretion. A new dendritic cell–specific member of the B7 family, B7-DC, was also found to bind PD-1 (224) and to efficiently costimulate T cell proliferation, as well as IL-2 and IFN-γ secretion. 4-1BB, an activating T cell costimulatory molecule, is engaged by 4-1BBL at the surface of dendritic cells (225). 4-1BB/4-1BBL signaling leads to increased CD8+ T cell activation. By contrast, OX40 ligation on T cells by OX40L on the dendritic cell is involved in the differentiation of CD4+ T cells (226). The in vivo relevance of dendritic cell maturation in T cell priming is particularily clear in the case of CD8+ T cell priming. Despite several examples of direct priming of CTL by dendritic cells in the absence of CD4+ T cells (227), several in vivo CD8+ T cell responses are dependent on CD4+ T cell help (228). In addition, CD4+ T cells may also convert induction of CD8+ T cell tolerance into priming, in cases of constitutive cross presentation of peripheral self-antigens (229). Concerning cross priming of OVA-specific CTLs after the injection of MHCmismatched OVA-loaded cells, Bennett et al. first demonstrated that CD4+ T cell help required antigen recognition on the same APC (228). It was then discovered that CD40 ligation on APCs by CD40L on CD4+ T cells was necessary and

4 Feb 2002

11:53

Annu. Rev. Immunol. 2002.20:621-667. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

640

AR

AR152-20.tex

AR152-20.SGM

LaTeX2e(2001/05/10)

P1: GJC

GUERMONPREZ ET AL.

sufficient to confer to APCs the ability to prime CTLs (230–232). CD40 ligation is sufficient to drive the maturation of dendritic cells (76) and confer on them the ability to prime CD8+ T cells (75). The licensing effect of CD40 ligation was also demonstrated upon injection of antigen-loaded dendritic cells (233). Moreover, CD40 ligation can also overcome the tolerogenic effect of optimal-length peptide administration, giving rise to a great interest in manipulating this pathway in the field of cancer vaccines (234). This mechanism allows a temporal dissociation of CD4+ and CD8+ antigen-specific T cell interaction with the dendritic cells: Once licensed by CD4+ T cells, dendritic cells become competent to prime CD8+ T cells. This model bypasses the requirement for an improbable simultaneous encounter of antigen-bearing dendritic cells with antigen-specific CD4+ and CD8+ T cells. CD4+ T cell help may also occur in a CD4+ T cell–dependent, CD40-independent manner (235), as in the case of some anti-viral responses (227), possibly by the ligation of other receptors belonging to the TNF-R family, like RANK/TRANCEL (see Do T Cells Regulate DC Half-Life?). Finally, dendritic cell licensing may occur in a CD4+ T cell–independent manner. Pathogen-derived products that drive dendritic cell maturation, such as LPS (75) or viruses (231), can also license dendritic cells to efficiently prime CTL responses. The expression of CD40L by some apoptotic bodies can deliver a licensing signal that bypasses the CD4+ T cell help for cross priming of antigens expressed by these apoptotic bodies (83). DO IMMATURE DENDRITIC CELLS DIRECTLY TOLERIZE T CELLS? In vitro studies show that recognition and phagocytosis by dendritic cells of apoptotic cells, which are normally produced by the constitutive renewal of tissues, do not necessarily induce dendritic cell maturation (40, 41). Induction of the chemokine receptor CCR7 upon phagocytosis of apoptotic bodies (236) suggests that, in vivo, some dendritic cells might reach lymph nodes in a non–fully mature state following chemoattraction toward MIP-3β and SLC gradients. In addition, basal migratory flux of dendritic cell from tissues to lymph nodes is observed in the absence of maturation stimuli. Such constitutive dendritic cell migration was recently observed from intestinal epithelium to the T cell area of mesenteric lymph nodes in gnotobiotic rats (45). It was proposed that T cell recognition of self-peptide/MHC complexes on non–fully mature dendritic cells could induce tolerance (237). Alternatively, antigen presentation by immature dendritic cells might induce differentiation of na¨ıve self-reactive T cells toward a suppressor/regulatory phenotype (238, 239). Indeed, repeated stimulations of allogenic human CD4+ T cells by immature dendritic cells induces IL-10-producing, weakly proliferative CTLA4+ T cells (240). These cells inhibit the mature dendritic cell–mediated Th1 T cell differentiation. In addition, injection of peptide-pulsed immature dendritic cells in healthy volunteers specifically suppressed CD8+ memory T cell responses, through the induction of nonlytic antigen-specific T cells secreting IL-10, low levels of IFN-γ , and no IL-4 (241). Quantification of antigen-specific T cells revealed that the loss

4 Feb 2002

11:53

AR

AR152-20.tex

AR152-20.SGM

LaTeX2e(2001/05/10)

P1: GJC

Annu. Rev. Immunol. 2002.20:621-667. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

ANTIGEN PRESENTATION BY DENDRITIC CELLS

641

of effector functions was not due to a decrease in the number of circulating T cells. A growing list of stimuli such as IL-10, glucocorticoids, vitamin D3, tumorderived products, or Plasmodium falciparum-infected erythrocytes, measles, or herpes virus infection appears to inhibit dendritic cell terminal differentiation at various stages, and in some cases, favors the induction of anergy in T cells or the priming of regulatory Tr1 cells (secreting IL-10, but not IFN-γ and IL-4) (219, 239). In vivo, the tolerogenicity of immature dendritic cells could lead to therapeutic applications because the injection of allogenic dendritic cells displaying an immature phenotype prolongs allograft survival (242). PLASTICITY OF THE MATURATION AND CD4+ T CELL POLARIZATION

It is now clear that the developmental origin of dendritic cells does not fully determine their capacity to polarize T cells. Other factors condition the outcome of T cell stimulation by dendritic cells: 1. Human monocyte–derived dendritic cells matured by either CD40L (243), IFN-γ LPS, oligo CpG nucleotides (244), or double-stranded RNA (106) produce high levels of IL-12 and induce Th1 cells. Nevertheless, stimuli like PGE2 may favor the differentiation of mature monocyte-derived dendritic cells priming Th2 cells (74). The concept of polarized maturation is now widely conforted by the growing list of factors that allow the generation of mature IL-12-nonsecreting dendritic cells, such as PGE2, β2-agonists, TGFβ, IL-1, TNF-α, and Fas engagement (74, 219). Moreover, certain dendritic cell costimulatory molecules, such as OX40L, can deliver a Th2 polarizing signal to CD4+ T cells (219). Importantly, human plasmacytoid dendritic cells mature and secrete high doses of IL-12 and IFN-α upon viral infection and prime Th1 differentiation (through the induction of IFN-γ in human CD4+ T cells). They correspond to the previously described natural interferon producing cells (210, 245–247). Finally, monocyte-derived and plasmacytoid dendritic cells regulate each other: IFN-α favors DC1 differentiation and inhibits IL-12 production, thus promoting Th2 responses, whereas IL-4 induces apoptosis of preDC2, thus favoring DC1differentiation and IL-12 production (219). 2. Pathogen-derived signals appear to play a crucial role in the regulation of cytokine secretion by dendritic cells. For example, Toxoplasma gondii extracts induce IL-12 production by murine CD8α + dendritic cells in vivo and sensitivity to CD40 stimulation. However, a prolonged stimulation by the extracts paralyses IL-12 production by dendritic cells (65, 248). Candida albicans, at the yeast stage, induces IL-12 production and Th1 protective responses in vitro and in vivo. At the hyphae stage, the same fungus induces dendritic cells to stimulate IL-4 secretion by T cells and no protective responses (31). 3. Tissue-specific environmental factors may participate in the phenotypic differentiation of dendritic cells. In mouse, CD11c+ dendritic cells purified

4 Feb 2002

11:53

642

AR

AR152-20.tex

AR152-20.SGM

LaTeX2e(2001/05/10)

P1: GJC

GUERMONPREZ ET AL.

from Peyer’s patches or lung, but not spleen, produce IL-4 and IL-10 and preferentially induce in vitro Th2 polarization (249, 250). 4. Dendritic cell/T cell ratio (251) or the duration of dendritic cell stimulation can also modulate the polarization of CD4+ T cells. Indeed, after prolonged stimulation with LPS, IL-12 production in response to CD40L (or LPS) is reduced. These exhausted dendritic cells prime mostly for Th2 and nonpolarized T cells expressing CCR7, also called central memory T cells (252).

Annu. Rev. Immunol. 2002.20:621-667. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

Do T Cells Regulate Dendritic Cell Half-Life? T cells receiving antigenic information and priming/tolerizing signals may, in turn, regulate the half-life of antigen-bearing dendritic cells by several mechanisms. This kind of regulation may crucially affect the dynamics of T cell priming in vivo. Indeed, CD4+ T cell commitment to proliferation requires about 10 h of sustained signaling in the presence of costimulation (253). In contrast, CD8+ T cells require only a short exposure to antigen and costimulation to be committed to proliferation. Strikingly, once committed, T cells proliferate and differentiate without the need for further stimulation (254). The half-life of antigen-bearing dendritic cells may thus represent a crucial parameter of T cell priming. Prolonging dendritic cell half-life by TRANCE treatment results in increased numbers of dendritic cells reaching the lymph nodes and increased T cell priming (255). Ex vivo antigen-loaded dendritic cells rapidly disappear from lymph nodes after antigen-specific interaction with na¨ıve CD4+ T cells (188). However, antigen-specific interaction between dendritic cells and CD4+ T cells also delays LPS-induced apoptosis of splenic dendritic cells (256). Concerning CD8+ T cells, the time course of acquisition of lytic effector function (more rapid for memory than for na¨ıve cells) parallels the kinetics of elimination of antigen-bearing dendritic cells (257, 258). If CTL-mediated elimination of dendritic cells limits the priming of further immune responses (257), it does not impede the restimulation of memory CTL responses (259). Dendritic cell half-life is regulated by T cell–derived signals through: (a) Cytokine receptors: DC2-primed Th2 cells secrete IL-4 and IL-10 that synergistically inhibits the proliferating effects of IL-3 on DC2, while IL-4 increases the maturation of monocyte-derived dendritic cells (221). (b) MHC class II molecules: Mature dendritic cells are highly susceptible to cell death induced by MHC class II engagement (260). (c) RANK/TRANCE-L: Ligation by RANK-R/TRANCE promotes survival in bone marrow–derived dendritic cells, increasing their T cell stimulatory ability (261, 262). This leads to increased survival in the lymph node of subcutaneously injected antigen-loaded dendritic cells, thus favoring the expansion of antigen-specific CD4+ T cells (255). (d) TNF-R: TNF-R2 signaling plays a crucial role in the survival of LCs (263) and bone marrow–derived dendritic cells and migration of LCs (264), whereas TNF-R1 is implicated in the NF-κB-dependent maturation

4 Feb 2002

11:53

AR

AR152-20.tex

AR152-20.SGM

LaTeX2e(2001/05/10)

P1: GJC

ANTIGEN PRESENTATION BY DENDRITIC CELLS

643

Annu. Rev. Immunol. 2002.20:621-667. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

process of human monocyte-derived dendritic cells (52) and murine bone marrow–derived dendritic cells (265). (e) Death signaling receptors: Dendritic cell maturation triggers resistance to Fas- or TRAIL-mediated death signals in correlation with increased expression of the caspase 8 inhibitory protein cFLIP (266, 267). Rescigno et al. described a Fas-dependent pathway for the induction of maturation in correlation with the constitutive expression of FLICE-inhibitory protein (FLIP) (77). Another study, however, showed induction of apoptosis by Fas ligand in immature dendritic cells (268). ( f ) The perforin/granzyme pathway: Although dendritic cells express the protease inhibitor 9 (269) and SPI-6 (270), a serpin family member, that both inhibit granzyme B, they can be eliminated by CTLs in vitro by the perforindependent pathway (258) and in vivo (257) by perforin-independent unknown mechanisms (258). It is interesting that Th1 but not Th2 induces SPI-6 expression in dendritic cells (270).

DENDRITIC CELL–BASED IMMUNOTHERAPY The analysis of the mechanisms underlying antigen presentation and T cell stimulation provided the rationale for developing a novel strategy for immunotherapy, based on the injection of antigen-pulsed dendritic cells. Several issues need to be considered to ensure optimal outcome of dendritic cell-based vaccination protocols, including (a) a good manufacturing procedure used to expand dendritic cells and (b) the source and formulation of antigen for appropriate peptide loading on MHC molecules. These issues are still open, and only an extensive understanding of dendritic cell fundamental biology will allow us to develop effective immunotherapy protocols. It is also important to consider, however, that this novel immunotherapy approach is endowed with novel caveats and risks.

Manipulation of Dendritic Cells for Immunotherapy EXPANDING DENDRITIC CELLS FOR THERAPEUTICAL USE Two alternative strategies are currently being explored for dendritic cell expansion in immunotherapy protocols: ex vivo and in vivo. Ex vivo, several cytokine cocktails generate immature or mature monocyte-derived dendritic cells or CD34+ precursor-derived dendritic cells from normal volunteers or patients. Some of these dendritic cell preparations are compatible with clinical grade utilization (6). Particular attention should be drawn toward the use of fetal calf serum (case report of anaphylaxis) (271). In vivo, Flt3L injected in healthy individuals elicits a profound increase in the numbers of immature dendritic cells and precursor dendritic cells in peripheral blood (both DC1 and DC2). Whether this dramatic increase in dendritic cell numbers in vivo enhances immune responses to vaccine antigens remains to be established (272).

4 Feb 2002

11:53

Annu. Rev. Immunol. 2002.20:621-667. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

644

AR

AR152-20.tex

AR152-20.SGM

LaTeX2e(2001/05/10)

P1: GJC

GUERMONPREZ ET AL. ANTIGEN DELIVERY TO DENDRITIC CELLS The ultimate goal is to use an antigenic formulation allowing a wide peptide repertoire presentation to both CD4+ and CD8+ T cells. Efficient and stable MHC class I and II presentation is obtained when antigens are synthesized endogenously by the dendritic cells. This is achieved in some malignant cells in patients with acute or chronic myelogenous leukemia, which, upon cytokine stimulation, undergo differentiation into APCs resembling dendritic cells that contain the entire repertoire of tumor antigens (273–275). For other diseases, the antigen must be delivered to dendritic cells ex vivo. Endogenous antigen expression may be obtained using bulk RNA prepared from the tumor (276). Genetic ex vivo manipulation of dendritic cells to express full-length antigens allows presentation of both MHC class I and II epitopes for any MHC haplotype and the possibility to include sequences encoding immunomodulators (6). The use of RNA, DNA, and substractive hybridization could allow for enrichment and fast amplification of tumor-specific RNA (6). Universal vaccination using mRNA encoding the polypeptide component of telomerase has been successful in eliciting tumor-specific CTLs in various tumor models, leading to effective and broad cross protection in tumor-bearing mice (277). Retroviral/lentiviral, adenoviral, and poxviral vectors have also been used to elicit effective tumor antigen- and viral antigen-specific CTL responses in vitro and in vivo (278). While retroviral vectors are suitable to transduce dividing CD34+ progenitor-derived dendritic cells (up to 70% efficiency), vaccinia or adenoviral vectors transduce monocyte-derived dendritic cells. However, one should consider the potential ability of the virus to interfere with the antigen presentation machinery of the infected dendritic cells, to affect their maturation, or to cause apoptosis (278). Regarding the problems associated with vector-specific immunity, primeboost strategies will have to be tested. Efficient nonviral transfection of dendritic cells using plasmid DNA expression constructs with cationic peptide CL22 elicit efficient antitumor responses in mice (279). Another approach to obtain endogenous antigen synthesis has been to fuse dendritic cells to tumor cells. Pioneer work by Gong et al. (280), showing that murine dendritic cells fused to colon cancer cells elicit tumor-specific CTLs in vitro and in vivo, has been extended to several other experimental systems, confirming the immunogenicity of such vaccines (281). Although it was postulated that dendritic cell-tumor fusion hybrids might allow tumor antigen to gain access to the dendritic cell cytosol for effective MHC class I presentation, the precise mechanisms of the immune responses leading to the clinical success of the allogeneic approach remain to be clarified. The alternative strategy to endogenous antigen synthesis is to provide antigens exogenously. Peptide-pulsed bone marrow dendritic cells generate antigen-specific T cell–mediated immune responses in mice, leading to tumor eradication (179, 282–284). Pulsing synthetic peptides derived from known tumor (285–287) or viral (288) antigen precursors onto human ex vivo–generated dendritic cells elicited peptide-specific CTL responses in vitro. These peptides, however, only reside on the dendritic cell surface for short periods of time (289, 290), are limited in use

4 Feb 2002

11:53

AR

AR152-20.tex

AR152-20.SGM

LaTeX2e(2001/05/10)

P1: GJC

Annu. Rev. Immunol. 2002.20:621-667. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

ANTIGEN PRESENTATION BY DENDRITIC CELLS

645

for patients bearing the appropriate MHC haplotypes, and were associated with antigen and/or MHC class I loss variants in vivo (291, 292). As mentioned in MHC Class I Restricted Antigen Presentation in Dendritic Cells, apoptotic cells also allow efficient antigen loading in dendritic cells. NouriShirazi et al. (118) and B´erard et al. (38) have shown that dendritic cells loaded with killed allogeneic melanoma or prostate cells prime na¨ıve T cells to differentiate into CTLs specific for a broad spectrum of shared tumor antigens and recognizing naturally processed antigens. We have shown that a novel tumor cell secretory compartment, exosomes, is a defined source of shared rejection tumor antigens released in membrane vesicles by living tumor cells. Tumor-derived exosomes contain whole native cytosolic and/or endosomal tumor antigens and constitutive hsps. Exosomes transfer tumor antigen to dendritic cells and induce peptide-specific, MHC class I–restricted cross presentation to T cell clones and in vitro tumor-specific CTL responses in patients’ lymphocytes (renal cell cancer and melanoma). Moreover, exosomes promote cross protecting antitumor effects in both syngeneic and allogeneic mouse tumor models, suggesting that they may transfer shared tumor antigens (120).

Potential Caveats of Dendritic Cell–Based Immunotherapy of Cancer Activated dendritic cells can break tumor-induced tolerance. Tumor antigens include tissue-specific antigens such as melanoma/melanocyte antigens and antigens widely expressed in peripheral normal tissues, albeit at low levels. To use such tumor antigens for immunotherapy, one should consider the risk of autoimmunity. For example, dendritic cells pulsed with a peptide from a natural melanocyte/melanoma tumor antigen could induce T cell–dependent autoimmune responses, leading to melanoma rejection and vitiligo due to autoimmunity at the vaccination site (194). In mice expressing LCMV antigen in the pancreas, efficient dendritic cell vaccination against a LCMV-expressing tumor led to overt diabetes (195). In humans, a link between dendritic cell–mediated cross presentation and autoimmunity was suggested by the analysis of patients with autoimmune paraneoplastic cerebellar degeneration (PCD), who have limited breast or ovarian cancer evolution, and an expanded population of anti-cdr2 CTLs (cdr2 is normally expressed in neurons and testis) (293). Dendritic cells from PCD patients phagocytosed apoptotic tumor lines expressing cdr2 and induced potent anti-cdr2 cytolysis from autologous T cells. In PCD, cross presentation of apoptotic tumor cells by dendritic cells might provide the initial stimulus for CTLs in vivo. Another potential problem is that, if cross presentation allows the priming of protective CTL responses against tumors or infectious agents, it could also result in the priming of CTLs specific for antigenic structures that are not presented by infected or tumoral target cells and, thus, are devoid of protective ability (294, 295). Conversely, dendritic cell processing of tumoral antigens by the immunoproteasome can impede the presentation of some epitopes otherwise presented in tumoral cells devoid of immunoproteasome (109).

4 Feb 2002

11:53

646

AR

AR152-20.tex

AR152-20.SGM

LaTeX2e(2001/05/10)

P1: GJC

GUERMONPREZ ET AL.

Annu. Rev. Immunol. 2002.20:621-667. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

Human Trials The first controlled evidence of the adjuvant role of mature dendritic cells in normal volunteers was brought up in 1999/2000 by Steinman’s group (296). While injection of unpulsed dendritic cells or antigen alone (KLH, TT, MP) failed to immunize, priming of CD4+ T cells to KLH, boosting of TT specific T cell immunity, and enhanced precursor frequency of MP-specific IFN-γ producing CTLs were shown following injection of a single dose of mature dendritic cells. When T cells were boosted in culture, there was an increase in MHC-specific tetramer-binding cells and CTL after dendritic cell vaccination. Responses to all three antigens peaked at 30–90 days after immunization and declined thereafter. Three volunteers were reinjected with mature dendritic cells pulsed with matrix protein alone and displayed a rapid boost in matrix protein-specific immunity (faster and greater magnitude of CD8+ T cell specific responses), with direct cytolytic activity in one case. Clinical trials using dendritic cell-based immunotherapy have been initiated or completed in melanoma, lymphoma, myeloma, prostate, and renal cancer patients (297). All published trials demonstrated safety. In most of these studies, immature monocyte-derived dendritic cells pulsed with tumor antigen peptides were used (239, 298–300). The recent observation (see Dendritic Cell Maturation and T Cell Priming) that injection of immature dendritic cells in human suppresses CTL responses (241) suggests, however, that mature dendritic cells could be a better antitumor adjuvant. A series of clinical trials based on the blood cell sorting of dendritic cells upfront (301) or following Flt3L therapy (so-called mature dendritic cells after two days in culture for antigen loading) was recently published (272, 302, 303). While Flt3L directly injected in metastatic patients was mediating dendritic cell expansion, no objective clinical response was seen (some enhanced DTH responses to common antigens and tumor dendritic cells infiltrates were described). Using monocyte-derived, peptide-pulsed dendritic cells fully matured under rigorous quality control parameters (GM-CSF/IL4 followed by IL1β, TNFα, PGE2, IL-6), two groups (304, 305) reported clinical and immunological responses in melanoma or renal cancer patients, specially in those with low tumor burden. CD34+-derived dendritic cells are also being used in clinical trials (306). A provocative report from Kugler et al. (307) recently described therapy of 17 renal cell carcinoma patients with tumor cell–dendritic cell hybrids (electrofusion). Dendritic cells were allogeneic, whereas tumor cells were autologous. The mean follow up was 13 months, and 4 complete responses, 2 partial responses, and 1 minor response were achieved on metastatic diseases. They also demonstrated recruitment of CD8+ T lymphocytes into tumor sites. Fusions and hybrid formation were documented only in 10%–15% of total reinjected cells. Safety and antigen presenting properties of allogeneic (HLA-A2, pulsed with Gal, Pol, Env peptides) or autologous dendritic cells were addressed in 6 HLAA2+, HIV infected patients. Both dendritic cell infusions were well tolerated, and in patients with normal CD4+ T cell counts, administration of these antigen-pulsed cells enhanced the immune response to HIV with no effect on the viral load (308).

4 Feb 2002

11:53

AR

AR152-20.tex

AR152-20.SGM

LaTeX2e(2001/05/10)

P1: GJC

Annu. Rev. Immunol. 2002.20:621-667. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

ANTIGEN PRESENTATION BY DENDRITIC CELLS

647

They further showed persistance of donor cells following adoptive transfer of allogeneic dendritic cells. Besides antigen-pulsed dendritic cells injection in patients, alternative strategies could rely on direct expansion of dendritic cells in patients or on targeting antigens to dendritic cells in vivo. Hsps could represent such a vehicle to deliver peptides to dendritic cells. Alternatively, we showed that, like tumor cells, murine dendritic cells secrete antigen presenting vesicles of endosomal origin, called exosomes (309). Dendritic cell–derived exosomes induce potent antitumor immune responses and eradication. They have a unique protein composition, distinct from other cellular compartments, including apoptotic microvesicles (310, 311). Peptide-loaded exosomes are efficiently presented by murine dendritic cells to CD4+ T cells in vitro and can stimulate na¨ıve T cells in vivo (C. Th´ery, L. Duban, P. Veron, O. Lantz, S. Amigorena, et al., in preparation). Immature human dendritic cells also secrete exosomes of similar protein composition and function (309). A good manufacturing process has been engineered allowing scale up of exosome production and purification with safety profile of the product in mice. A clinical trial has recently been launched, aimed at vaccinating subcutaneously metastatic melanoma and inoperable lung cancer patients with autologous dendritic cell–derived exosomes pulsed with Mage3 class I and II peptides. Ideal vaccination processes should be cell free, off-shelf reagents, allow manufacturing in large scale, and use immunorelevant tumor antigens regardless of patient HLA haplotype. We are, obviously, far from this goal. In spite of encouraging preliminary clinical results in pilot studies and phase I/II trials (demonstrating feasibility and safety), dendritic cell–based immunotherapy awaits firm proof of efficacy. In particular, correlations between enhanced specific CTL precursor frequency and clinical responses have not been reported yet. ACKNOWLEDGMENTS P.G. was supported by Association pour la Recherche contre le Cancer, J.V. by Ligue Nationale contre le Cancer, C.T. by Fondation pour la Recherche M´edicale. Due to space limitation, we could only cite a fraction of the published work, which does not undermine the great value of uncited studies. Visit the Annual Reviews home page at www.annualreviews.org

LITERATURE CITED 1. Banchereau J, Steinman RM. 1998. Dendritic cells and the control of immunity. Nature 392:245–52 2. Banchereau J, Briere F, Caux C, Davoust J, Lebecque S, Liu YJ, Pulendran B, Palucka K. 2000. Immunobiology of den-

dritic cells. Annu. Rev. Immunol. 18:767– 811 3. Bell D, Young JW, Banchereau J. 1999. Dendritic cells. Adv. Immunol. 72:255– 324 4. Thery C, Amigorena S. 2001. The cell

4 Feb 2002

11:53

648

5.

6.

Annu. Rev. Immunol. 2002.20:621-667. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

7.

8.

9.

10.

11.

12.

13.

AR

AR152-20.tex

AR152-20.SGM

LaTeX2e(2001/05/10)

P1: GJC

GUERMONPREZ ET AL. biology of antigen presentation in dendritic cells. Curr. Opin. Immunol. 13:45– 51 Gallucci S, Matzinger P. 2001. Danger signals: SOS to the immune system. Curr. Opin. Immunol. 13:114–19 Nestle FO, Banchereau J, Hart D. 2001. Dendritic cells: on the move from bench to bedside. Nat. Med. 7:761–65 Steinman RM, Swanson J. 1995. The endocytic activity of dendritic cells. J. Exp. Med. 182:283–88 Slepnev VI, De Camilli P. 2000. Accessory factors in clathrin-dependent synaptic vesicle endocytosis. Nat. Neurosci. Rev. 1:161–72 Fanger NA, Wardwell K, Shen L, Tedder TF, Guyre PM. 1996. Type I (CD64) and type II (CD32) Fc gamma receptormediated phagocytosis by human blood dendritic cells. J. Immunol. 157:541–48 Esposito-Farese ME, Sautes C, de la Salle H, Latour S, Bieber T, de la Salle C, Ohlmann P, Fridman WH, Cazenave JP, Teillaud JL, et al. 1995. Membrane and soluble Fc gamma RII/III modulate the antigen-presenting capacity of murine dendritic epidermal Langerhans cells for IgG-complexed antigens. J. Immunol. 155:1725–36 Geissmann F, Launay P, Pasquier B, Lepelletier Y, Leborgne M, Lehuen A, Brousse N, Monteiro RC. 2001. A subset of human dendritic cells expresses IgA Fc receptor (CD89), which mediates internalization and activation upon crosslinking by IgA complexes. J. Immunol. 166:346–52 Jurgens M, Wollenberg A, Hanau D, de la Salle H, Bieber T. 1995. Activation of human epidermal Langerhans cells by engagement of the high affinity receptor for IgE, Fc epsilon RI. J. Immunol. 155:5184– 89 Zhu X, Meng G, Dickinson BL, Li X, Mizoguchi E, Miao L, Wang Y, Robert C, Wu B, Smith PD, Lencer WI, Blumberg RS. 2001. MHC class I-related neonatal

14.

15.

16.

17.

18.

19.

20.

21.

Fc receptor for IgG is functionally expressed in monocytes, intestinal macrophages, and dendritic cells. J. Immunol. 166:3266–76 Cella M, Dohring C, Samaridis J, Dessing M, Brockhaus M, Lanzavecchia A, Colonna M. 1997. A novel inhibitory receptor (ILT3) expressed on monocytes, macrophages, and dendritic cells involved in antigen processing. J. Exp. Med. 185: 1743–51 Reis e Sousa C, Stahl PD, Austyn JM. 1993. Phagocytosis of antigens by Langerhans cells in vitro. J. Exp. Med. 178: 509–19 Castellino F, Boucher PE, Eichelberg K, Mayhew M, Rothman JE, Houghton AN, Germain RN. 2000. Receptor-mediated uptake of antigen/heat shock protein complexes results in major histocompatibility complex class I antigen presentation via two distinct processing pathways. J. Exp. Med. 191:1957–64 Arnold-Schild D, Hanau D, Spehner D, Schmid C, Rammensee HG, de la Salle H, Schild H. 1999. Cutting edge: receptormediated endocytosis of heat shock proteins by professional antigen-presenting cells. J. Immunol. 162:3757–60 Basu S, Binder RJ, Ramalingam T, Srivastava PK. 2001. CD91 is a common receptor for heat shock proteins gp96, hsp90, hsp70, and calreticulin. Immunity 14:303–13 Platt N, da Silva RP, Gordon S. 1998. Recognizing death: the phagocytosis of apoptotic cells. Trends Cell Biol. 8:365– 72 Nakamura K, Funakoshi H, Miyamoto K, Tokunaga F, Nakamura T. 2001. Molecular cloning and functional characterization of a human scavenger receptor with C-type lectin (SRCL), a novel member of a scavenger receptor family. Biochem. Biophys. Res. Commun. 280:1028–35 Sallusto F, Cella M, Danieli C, Lanzavecchia A. 1995. Dendritic cells use macropinocytosis and the mannose receptor

4 Feb 2002

11:53

AR

AR152-20.tex

AR152-20.SGM

LaTeX2e(2001/05/10)

P1: GJC

ANTIGEN PRESENTATION BY DENDRITIC CELLS

Annu. Rev. Immunol. 2002.20:621-667. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

22.

23.

24.

25.

26.

27.

28.

to concentrate macromolecules in the major histocompatibility complex class II compartment: downregulation by cytokines and bacterial products. J. Exp. Med. 182:389–400 Jiang W, Swiggard WJ, Heufler C, Peng M, Mirza A, Steinman RM, Nussenzweig MC. 1995. The receptor DEC-205 expressed by dendritic cells and thymic epithelial cells is involved in antigen processing. Nature 375:151–55 Ezekowitz RA, Williams DJ, Koziel H, Armstrong MY, Warner A, Richards FF, Rose RM. 1991. Uptake of Pneumocystis carinii mediated by the macrophage mannose receptor. Nature 351:155–58 Prigozy TI, Sieling PA, Clemens D, Stewart PL, Behar SM, Porcelli SA, Brenner MB, Modlin RL, Kronenberg M. 1997. The mannose receptor delivers lipoglycan antigens to endosomes for presentation to T cells by CD1b molecules. Immunity 6:187–97 Dong X, Storkus WJ, Salter RD. 1999. Binding and uptake of agalactosyl IgG by mannose receptor on macrophages and dendritic cells. J. Immunol. 163:5427–34 Mahnke K, Guo M, Lee S, Sepulveda H, Swain SL, Nussenzweig M, Steinman RM. 2000. The dendritic cell receptor for endocytosis, DEC-205, can recycle and enhance antigen presentation via major histocompatibility complex class IIpositive lysosomal compartments. J. Cell Biol. 151:673–84 Kato M, Neil TK, Fearnley DB, McLellan AD, Vuckovic S, Hart DN. 2000. Expression of multilectin receptors and comparative FITC-dextran uptake by human dendritic cells. Int. Immunol. 12:1511–19 Valladeau J, Ravel O, Dezutter-Dambuyant C, Moore K, Kleijmeer M, Liu Y, Duvert-Frances V, Vincent C, Schmitt D, Davoust J, Caux C, Lebecque S, Saeland S. 2000. Langerin, a novel C-type lectin specific to Langerhans cells, is an endocytic receptor that induces the formation of Birbeck granules. Immunity 12:71–81

649

29. Geijtenbeek TB, Krooshoop DJ, Bleijs DA, van Vliet SJ, van Duijnhoven GC, Grabovsky V, Alon R, Figdor CG, van Kooyk Y. 2000. DC-SIGN-ICAM-2 interaction mediates dendritic cell trafficking. Nat. Immunol. 1:353–57 30. Rescigno M, Urbano M, Valzasina B, Francolini M, Rotta G, Bonasio R, Granucci F, Kraehenbuhl JP, RicciardiCastagnoli P. 2001. Dendritic cells express tight junction proteins and penetrate gut epithelial monolayers to sample bacteria. Nat. Immunol. 2:361–67 31. d’Ostiani CF, Del Sero G, Bacci A, Montagnoli C, Spreca A, Mencacci A, Ricciardi-Castagnoli P, Romani L. 2000. Dendritic cells discriminate between yeasts and hyphae of the fungus Candida albicans. Implications for initiation of T helper cell immunity in vitro and in vivo. J. Exp. Med. 191:1661– 74 32. Gorak PM, Engwerda CR, Kaye PM. 1998. Dendritic cells, but not macrophages, produce IL-12 immediately following Leishmania donovani infection. Eur. J. Immunol. 28:687–95 33. Konecny P, Stagg AJ, Jebbari H, English N, Davidson RN, Knight SC. 1999. Murine dendritic cells internalize Leishmania major promastigotes, produce IL12 p40 and stimulate primary T cell proliferation in vitro. Eur. J. Immunol. 29:1803– 11 34. Subklewe M, Paludan C, Tsang M, Mahnke K, Steinman R, Munz C. 2001. Dendritic cells cross-present latency gene products from Epstein-Barr virustransformed B cells and expand tumorreactive CD8(+) killer T cells. J. Exp. Med. 193:405–12 35. Albert ML, Sauter B, Bhardwaj N. 1998. Dendritic cells acquire antigen from apoptotic cells and induce class I-restricted CTLs. Nature 392:86–89 36. Shaif-Muthana M, McIntyre C, Sisley K, Rennie I, Murray A. 2000. Dead or alive: immunogenicity of human melanoma

4 Feb 2002

11:53

650

37.

Annu. Rev. Immunol. 2002.20:621-667. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

38.

39.

40.

41.

42.

43.

AR

AR152-20.tex

AR152-20.SGM

LaTeX2e(2001/05/10)

P1: GJC

GUERMONPREZ ET AL. cells when presented by dendritic cells. Cancer Res. 60:6441–47 Russo V, Tanzarella S, Dalerba P, Rigatti D, Rovere P, Villa A, Bordignon C, Traversari C. 2000. Dendritic cells acquire the MAGE-3 human tumor antigen from apoptotic cells and induce a class Irestricted T cell response. Proc. Natl. Acad. Sci. USA 97:2185–90 Berard F, Blanco P, Davoust J, NeidhartBerard EM, Nouri-Shirazi M, Taquet N, Rimoldi D, Cerottini JC, Banchereau J, Palucka AK. 2000. Cross-priming of naive CD8 T cells against melanoma antigens using dendritic cells loaded with killed allogeneic melanoma cells. J. Exp. Med. 192:1535–44 Hoffmann TK, Meidenbauer N, Dworacki G, Kanaya H, Whiteside TL. 2000. Generation of tumor-specific T-lymphocytes by cross-priming with human dendritic cells ingesting apoptotic tumor cells. Cancer Res. 60:3542–49 Sauter B, Albert ML, Francisco L, Larsson M, Somersan S, Bhardwaj N. 2000. Consequences of cell death: exposure to necrotic tumor cells, but not primary tissue cells or apoptotic cells, induces the maturation of immunostimulatory dendritic cells. J. Exp. Med. 191:423– 34 Gallucci S, Lolkema M, Matzinger P. 1999. Natural adjuvants: endogenous activators of dendritic cells. Nat. Med. 5: 1249–55 Inaba K, Turley S, Yamaide F, Iyoda T, Mahnke K, Inaba M, Pack M, Subklewe M, Sauter B, Sheff D, Albert M, Bhardwaj N, Mellman I, Steinman RM. 1998. Efficient presentation of phagocytosed cellular fragments on the major histocompatibility complex class II products of dendritic cells. J. Exp. Med. 188:2163– 73 Yrlid BU, Wick MJ. 2000. Salmonellainduced apoptosis of infected macrophages results in presentation of a bacteria-encoded antigen after uptake by

44.

45.

46.

47.

48.

49.

50.

51.

52.

bystander dendritic cells. J. Exp. Med. 191:613–24 Parr MB, Kepple L, Parr EL. 1991. Langerhans cells phagocytose vaginal epithelial cells undergoing apoptosis during the murine estrous cycle. Biol. Reprod. 45:252–60 Huang FP, Platt N, Wykes M, Major JR, Powell TJ, Jenkins CD, MacPherson GG. 2000. A discrete subpopulation of dendritic cells transports apoptotic intestinal epithelial cells to T cell areas of mesenteric lymph nodes. J. Exp. Med. 191:435– 44 Savill J, Fadok V. 2000. Corpse clearance defines the meaning of cell death. Nature 407:784–88 Fadok VA, Bratton DL, Rose DM, Pearson A, Ezekewitz RA, Henson PM. 2000. A receptor for phosphatidylserine-specific clearance of apoptotic cells. Nature 405:85–90 Scott RS, McMahon EJ, Pop SM, Reap EA, Caricchio R, Cohen PL, Earp HS, Matsushima GK. 2001. Phagocytosis and clearance of apoptotic cells is mediated by MER. Nature 411:207–11 Albert ML, Kim JI, Birge RB. 2000. alphavbeta5 integrin recruits the CrkIIDock180-rac1 complex for phagocytosis of apoptotic cells. Nat. Cell Biol. 2:899– 905 Albert ML, Pearce SF, Francisco LM, Sauter B, Roy P, Silverstein RL, Bhardwaj N. 1998. Immature dendritic cells phagocytose apoptotic cells via alphavbeta5 and CD36, and cross-present antigens to cytotoxic T lymphocytes. J. Exp. Med. 188:1359–68 Rubartelli A, Poggi A, Zocchi MR. 1997. The selective engulfment of apoptotic bodies by dendritic cells is mediated by the alpha(v)beta3 integrin and requires intracellular and extracellular calcium. Eur. J. Immunol. 27:1893–900 Sallusto F, Lanzavecchia A. 1994. Efficient presentation of soluble antigen by cultured human dendritic cells is

4 Feb 2002

11:53

AR

AR152-20.tex

AR152-20.SGM

LaTeX2e(2001/05/10)

P1: GJC

ANTIGEN PRESENTATION BY DENDRITIC CELLS

53.

Annu. Rev. Immunol. 2002.20:621-667. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

54.

55.

56.

57.

58.

59.

60.

maintained by granulocyte/macrophage colony-stimulating factor plus interleukin 4 and downregulated by tumor necrosis factor alpha. J. Exp. Med. 179:1109–18 West MA, Prescott AR, Eskelinen EL, Ridley AJ, Watts C. 2000. Rac is required for constitutive macropinocytosis by dendritic cells but does not control its downregulation. Curr. Biol. 10:839–48 Garrett WS, Chen LM, Kroschewski R, Ebersold M, Turley S, Trombetta S, Galan JE, Mellman I. 2000. Developmental control of endocytosis in dendritic cells by Cdc42. Cell 102:325–34 Marie JC, Kehren J, Trescol-Biemont MC, Evlashev A, Valentin H, Walzer T, Tedone R, Loveland B, Nicolas JF, Rabourdin-Combe C, Horvat B. 2001. Mechanism of measles virus-induced suppression of inflammatory immune responses. Immunity 14:69–79 Werling D, Hope JC, Chaplin P, Collins RA, Taylor G, Howard CJ. 1999. Involvement of caveolae in the uptake of respiratory syncytial virus antigen by dendritic cells. J. Leuk. Biol. 66:50–58 Blank C, Fuchs H, Rappersberger K, Rollinghoff M, Moll H. 1993. Parasitism of epidermal Langerhans cells in experimental cutaneous leishmaniasis with Leishmania major. J. Infect. Dis. 167:418–25 Gildea LA, Morris RE, Newman SL. 2001. Histoplasma capsulatum yeasts are phagocytosed via very late antigen-5, killed, and processed for antigen presentation by human dendritic cells. J. Immunol. 166:1049–56 Cameron P, Pope M, Granelli-Piperno A, Steinman RM. 1996. Dendritic cells and the replication of HIV-1. J. Leuk. Biol. 59:158–71 Patterson S, Rae A, Hockey N, Gilmour J, Gotch F. 2001. Plasmacytoid dendritic cells are highly susceptible to human immunodeficiency virus type 1 infection and release infectious virus. J. Virol. 75:6710– 13

651

61. Zaitseva M, Blauvelt A, Lee S, Lapham CK, Klaus-Kovtun V, Mostowski H, Manischewitz J, Golding H. 1997. Expression and function of CCR5 and CXCR4 on human Langerhans cells and macrophages: implications for HIV primary infection. Nat. Med. 3:1369–75 62. Geijtenbeek TB, Kwon DS, Torensma R, van Vliet SJ, van Duijnhoven GC, Middel J, Cornelissen IL, Nottet HS, KewalRamani VN, Littman DR, Figdor CG, van Kooyk Y. 2000. DC-SIGN, a dendritic cell-specific HIV-1-binding protein that enhances trans-infection of T cells. Cell 100:587–97 63. Geijtenbeek TB, Torensma R, van Vliet SJ, van Duijnhoven GC, Adema GJ, van Kooyk Y, Figdor CG. 2000. Identification of DC-SIGN, a novel dendritic cell-specific ICAM-3 receptor that supports primary immune responses. Cell 100:575–85 64. Medzhitov R, Janeway C Jr. 2000. Innate immunity. N. Engl. J. Med. 343:338–44 65. Reis e Sousa C. 2001. Dendritic cells as sensors of infection. Immunity 14:495–98 66. Aderem A, Ulevitch RJ. 2000. Toll-like receptors in the induction of the innate immune response. Nature 406:782–87 67. Visintin A, Mazzoni A, Spitzer JH, Wyllie DH, Dower SK, Segal DM. 2001. Regulation of Toll-like receptors in human monocytes and dendritic cells. J. Immunol. 166:249–55 68. Muzio M, Bosisio D, Polentarutti N, D’Amico G, Stoppacciaro A, Mancinelli R, van’t Veer C, Penton-Rol G, Ruco LP, Allavena P, Mantovani A. 2000. Differential expression and regulation of toll-like receptors (TLR) in human leukocytes: selective expression of TLR3 in dendritic cells. J. Immunol. 164:5998–6004 69. Michelsen KS, Aicher A, Mohaupt M, Hartung T, Dimmeler S, Kirschning CJ, Schumann RR. 2001. The role of toll-like receptors (TLRs) in bacteria-induced maturation of murine dendritic cells (DCS). Peptidoglycan and lipoteichoic acid are

4 Feb 2002

11:53

652

70.

Annu. Rev. Immunol. 2002.20:621-667. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

71.

72.

73.

74.

75.

76.

AR

AR152-20.tex

AR152-20.SGM

LaTeX2e(2001/05/10)

P1: GJC

GUERMONPREZ ET AL. inducers of dendritic cell maturation and require TLR2. J. Biol. Chem. 276:25680– 86 Hertz C, Kiertscher S, Godowski P, Bouis D, Norgard M, Roth M, Modlin R. 2001. Microbial lipopeptides stimulate dendritic cell maturation via Toll-like receptor 2. J. Immunol. 166:2444–50 Nishiguchi M, Matsumoto M, Takao T, Hoshino M, Shimonishi Y, Tsuji S, Begum NA, Takeuchi O, Akira S, Toyoshima K, Seya T. 2001. Mycoplasma fermentans lipoprotein M161Ag-induced cell activation is mediated by Toll-like receptor 2: role of N-terminal hydrophobic portion in its multiple functions. J. Immunol. 166:2610–16 Jeannin P, Renno T, Goetsch L, Miconnet I, Aubry JP, Delneste Y, Herbault N, Baussant T, Magistrelli G, Soulas C, Romero P, Cerottini JC, Bonnefoy JY. 2000. OmpA targets dendritic cells, induces their maturation and delivers antigen into the MHC class I presentation pathway. Nat. Immunol. 1:502–9 Underhill DM, Ozinsky A, Hajjar AM, Stevens A, Wilson CB, Bassetti M, Aderem A. 1999. The Toll-like receptor 2 is recruited to macrophage phagosomes and discriminates between pathogens. Nature 401:811–15 Kalinski P, Hilkens CM, Wierenga EA, Kapsenberg ML. 1999. T-cell priming by type-1 and type-2 polarized dendritic cells: the concept of a third signal. Immunol. Today 20:561–67 Schuurhuis DH, Laban S, Toes RE, Ricciardi-Castagnoli P, Kleijmeer MJ, van der Voort EI, Rea D, Offringa R, Geuze HJ, Melief CJ, Ossendorp F. 2000. Immature dendritic cells acquire CD8(+) cytotoxic T lymphocyte priming capacity upon activation by T helper cellindependent or -dependent stimuli. J. Exp. Med. 192:145–50 Caux C, Massacrier C, Vanbervliet B, Dubois B, Van Kooten C, Durand I, Banchereau J. 1994. Activation of human

77.

78.

79.

80.

81.

82.

83.

dendritic cells through CD40 crosslinking. J. Exp. Med. 180:1263–72 Rescigno M, Piguet V, Valzasina B, Lens S, Zubler R, French L, Kindler V, Tschopp J, Ricciardi-Castagnoli P. 2000. Fas engagement induces the maturation of dendritic cells (DCs), the release of interleukin (IL)-1beta, and the production of interferon gamma in the absence of IL-12 during DC-T cell cognate interaction: a new role for Fas ligand in inflammatory responses. J. Exp. Med. 192:1661–68 Ohshima Y, Tanaka Y, Tozawa H, Takahashi Y, Maliszewski C, Delespesse G. 1997. Expression and function of OX40 ligand on human dendritic cells. J. Immunol. 159:3838–48 Regnault A, Lankar D, Lacabanne V, Rodriguez A, Thery C, Rescigno M, Saito T, Verbeek S, Bonnerot C, Ricciardi-Castagnoli P, Amigorena S. 1999. Fcgamma receptor-mediated induction of dendritic cell maturation and major histocompatibility complex class I-restricted antigen presentation after immune complex internalization. J. Exp. Med. 189:371–80 Shi Y, Zheng W, Rock KL. 2000. Cell injury releases endogenous adjuvants that stimulate cytotoxic T cell responses. Proc. Natl. Acad. Sci. USA 97:14590–95 Rovere P, Vallinoto C, Bondanza A, Crosti MC, Rescigno M, Ricciardi-Castagnoli P, Rugarli C, Manfredi AA. 1998. Bystander apoptosis triggers dendritic cell maturation and antigen-presenting function. J. Immunol. 161:4467–71 Salio M, Cerundolo V, Lanzavecchia A. 2000. Dendritic cell maturation is induced by mycoplasma infection but not by necrotic cells. Eur. J. Immunol. 30:705–8 Propato A, Cutrona G, Francavilla V, Ulivi M, Schiaffella E, Landt O, Dunbar R, Cerundolo V, Ferrarini M, Barnaba V. 2001. Apoptotic cells overexpress vinculin and induce vinculin-specific cytotoxic T-cell cross-priming. Nat. Med. 7:807–13

4 Feb 2002

11:53

AR

AR152-20.tex

AR152-20.SGM

LaTeX2e(2001/05/10)

P1: GJC

Annu. Rev. Immunol. 2002.20:621-667. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

ANTIGEN PRESENTATION BY DENDRITIC CELLS 84. Basu S, Binder RJ, Suto R, Anderson KM, Srivastava PK. 2000. Necrotic but not apoptotic cell death releases heat shock proteins, which deliver a partial maturation signal to dendritic cells and activate the NF-kappa B pathway. Int. Immunol. 12:1539–46 85. Singh-Jasuja H, Scherer HU, Hilf N, Arnold-Schild D, Rammensee HG, Toes RE, Schild H. 2000. The heat shock protein gp96 induces maturation of dendritic cells and down-regulation of its receptor. Eur. J. Immunol. 30:2211–15 86. Vabulas RM, Ahmad-Nejad P, da Costa C, Miethke T, Kirschning CJ, Hacker H, Wagner H. 2001. Endocytosed heat shock protein 60s use TLR2 and TLR4 to activate the toll/interleukin-1 receptor signaling pathway in innate immune cells. J. Biol. Chem. 276:31,332–39 87. Ohashi K, Burkart V, Flohe S, Kolb H. 2000. Cutting edge: heat shock protein 60 is a putative endogenous ligand of the toll-like receptor-4 complex. J. Immunol. 164:558–61 88. Asea A, Kraeft SK, Kurt-Jones EA, Stevenson MA, Chen LB, Finberg RW, Koo GC, Calderwood SK. 2000. HSP70 stimulates cytokine production through a CD14-dependant pathway, demonstrating its dual role as a chaperone and cytokine. Nat. Med. 6:435–42 89. Kol A, Lichtman AH, Finberg RW, Libby P, Kurt-Jones EA. 2000. Cutting edge: heat shock protein (HSP) 60 activates the innate immune response: CD14 is an essential receptor for HSP60 activation of mononuclear cells. J. Immunol. 164:13– 17 90. Geissmann F, Revy P, Regnault A, Lepelletier Y, Dy M, Brousse N, Amigorena S, Hermine O, Durandy A. 1999. TGF-beta 1 prevents the noncognate maturation of human dendritic Langerhans cells. J. Immunol. 162:4567–75 91. Rescigno M, Martino M, Sutherland CL, Gold MR, Ricciardi-Castagnoli P. 1998. Dendritic cell survival and maturation are

92.

93.

94.

95.

96.

97.

98.

99.

653

regulated by different signaling pathways. J. Exp. Med. 188:2175–80 Ardeshna KM, Pizzey AR, Devereux S, Khwaja A. 2000. The PI3 kinase, p38 SAP kinase, and NF-kappaB signal transduction pathways are involved in the survival and maturation of lipopolysaccharidestimulated human monocyte-derived dendritic cells. Blood 96:1039–46 Vidalain PO, Azocar O, Servet-Delprat C, Rabourdin-Combe C, Gerlier D, Manie S. 2000. CD40 signaling in human dendritic cells is initiated within membrane rafts. EMBO J. 19:3304–13 Aicher A, Shu GL, Magaletti D, Mulvania T, Pezzutto A, Craxton A, Clark EA. 1999. Differential role for p38 mitogenactivated protein kinase in regulating CD40-induced gene expression in dendritic cells and B cells. J. Immunol. 163: 5786–95 Lu B, Yu H, Chow C, Li B, Zheng W, Davis RJ, Flavell RA. 2001. GADD45 gamma mediates the activation of the p38 and JNK MAP kinase pathways and cytokine production in effector TH1 cells. Immunity 14:583–90 Lyakh LA, Koski GK, Telford W, Gress RE, Cohen PA, Rice NR. 2000. Bacterial lipopolysaccharide, TNF-alpha, and calcium ionophore under serum-free conditions promote rapid dendritic cell-like differentiation in CD14+ monocytes through distinct pathways that activate NK-kappa B. J. Immunol. 165:3647–55 Kaisho T, Akira S. 2001. Dendriticcell function in Toll-like receptor- and MyD88-knockout mice. Trends Immunol. 22:78–83 Muzio M, Natoli G, Saccani S, Levrero M, Mantovani A. 1998. The human toll signaling pathway: divergence of nuclear factor kappaB and JNK/SAPK activation upstream of tumor necrosis factor receptor-associated factor 6 (TRAF6). J. Exp. Med. 187:2097–101 Kaisho T, Takeuchi O, Kawai T, Hoshino K, Akira S. 2001. Endotoxin-induced

4 Feb 2002

11:53

654

100.

Annu. Rev. Immunol. 2002.20:621-667. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

101.

102.

103.

104.

105.

106.

107.

108.

AR

AR152-20.tex

AR152-20.SGM

LaTeX2e(2001/05/10)

P1: GJC

GUERMONPREZ ET AL. maturation of myd88-deficient dendritic cells. J. Immunol. 166:5688–94 Sallusto F, Lanzavecchia A. 2000. Understanding dendritic cell and T-lymphocyte traffic through the analysis of chemokine receptor expression. Immunol. Rev. 177: 134–40 Randolph GJ, Beaulieu S, Lebecque S, Steinman RM, Muller WA. 1998. Differentiation of monocytes into dendritic cells in a model of transendothelial trafficking. Science 282:480–83 Bevan MJ. 1976. Cross-priming for a secondary cytotoxic response to minor H antigens with H-2 congenic cells which do not cross-react in the cytotoxic assay. J. Exp. Med. 143:1283–88 Cresswell P, Bangia N, Dick T, Diedrich G. 1999. The nature of the MHC class I peptide loading complex. Immunol. Rev. 172:21–28 Rescigno M, Citterio S, Thery C, Rittig M, Medaglini D, Pozzi G, Amigorena S, Ricciardi-Castagnoli P. 1998. Bacteriainduced neo-biosynthesis, stabilization, and surface expression of functional class I molecules in mouse dendritic cells. Proc. Natl. Acad. Sci. USA 95:5229–34 Cella M, Engering A, Pinet V, Pieters J, Lanzavecchia A. 1997. Inflammatory stimuli induce accumulation of MHC class II complexes on dendritic cells. Nature 388:782–87 Cella M, Salio M, Sakakibara Y, Langen H, Julkunen I, Lanzavecchia A. 1999. Maturation, activation, and protection of dendritic cells induced by doublestranded RNA. J. Exp. Med. 189:821–29 Macagno A, Gilliet M, Sallusto F, Lanzavecchia A, Nestle FO, Groettrup M. 1999. Dendritic cells up-regulate immunoproteasomes and the proteasome regulator PA28 during maturation. Eur. J. Immunol. 29:4037–42 MacAry PA, Lindsay M, Scott MA, Craig JI, Luzio JP, Lehner PJ. 2001. Mobilization of MHC class I molecules from late endosomes to the cell surface

109.

110.

111.

112.

113.

114.

115.

following activation of CD34-derived human Langerhans cells. Proc. Natl. Acad. Sci. USA 98:3982–87 Morel S, Levy F, Burlet-Schiltz O, Brasseur F, Probst-Kepper M, Peitrequin AL, Monsarrat B, Van Velthoven R, Cerottini JC, Boon T, Gairin JE, Van den Eynde BJ. 2000. Processing of some antigens by the standard proteasome but not by the immunoproteasome results in poor presentation by dendritic cells. Immunity 12:107–17 Bates EE, Ravel O, Dieu MC, Ho S, Guret C, Bridon JM, Ait-Yahia S, Briere F, Caux C, Banchereau J, Lebecque S. 1997. Identification and analysis of a novel member of the ubiquitin family expressed in dendritic cells and mature B cells. Eur. J. Immunol. 27:2471–77 Norbury CC, Chambers BJ, Prescott AR, Ljunggren HG, Watts C. 1997. Constitutive macropinocytosis allows TAP-dependent major histocompatibility complex class I presentation of exogenous soluble antigen by bone marrow-derived dendritic cells. Eur. J. Immunol. 27:280–88 Kovacsovics-Bankowski M, Clark K, Benacerraf B, Rock KL. 1993. Efficient major histocompatibility complex class I presentation of exogenous antigen upon phagocytosis by macrophages. Proc. Natl. Acad. Sci. USA 90:4942–46 Shen Z, Reznikoff G, Dranoff G, Rock KL. 1997. Cloned dendritic cells can present exogenous antigens on both MHC class I and class II molecules. J. Immunol. 158:2723–30 Reis e Sousa C, Germain RN. 1995. Major histocompatibility complex class I presentation of peptides derived from soluble exogenous antigen by a subset of cells engaged in phagocytosis. J. Exp. Med. 182:841–51 Oh YK, Harding CV, Swanson JA. 1997. The efficiency of antigen delivery from macrophage phagosomes into cytoplasm for MHC class I-restricted antigen presentation. Vaccine 15:511–18

4 Feb 2002

11:53

AR

AR152-20.tex

AR152-20.SGM

LaTeX2e(2001/05/10)

P1: GJC

Annu. Rev. Immunol. 2002.20:621-667. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

ANTIGEN PRESENTATION BY DENDRITIC CELLS 116. Wick MJ, Ljunggren HG. 1999. Processing of bacterial antigens for peptide presentation on MHC class I molecules. Immunol. Rev. 172:153–62 117. Arrode G, Boccaccio C, Lule J, Allart S, Moinard N, Abastado JP, Alam A, Davrinche C. 2000. Incoming human cytomegalovirus pp65 (UL83) contained in apoptotic infected fibroblasts is crosspresented to CD8(+) T cells by dendritic cells. J. Virol. 74:10018–24 118. Nouri-Shirazi M, Banchereau J, Bell D, Burkeholder S, Kraus ET, Davoust J, Palucka KA. 2000. Dendritic cells capture killed tumor cells and present their antigens to elicit tumor-specific immune responses. J. Immunol. 165:3797–803 119. Fields RC, Shimizu K, Mule JJ. 1998. Murine dendritic cells pulsed with whole tumor lysates mediate potent antitumor immune responses in vitro and in vivo. Proc. Natl. Acad. Sci. USA 95:9482–87 120. Wolfers J, Lozier A, Raposo G, Regnault A, Thery C, Masurier C, Flament C, Pouzieux S, Faure F, Tursz T, Angevin E, Amigorena S, Zitvogel L. 2001. Tumorderived exosomes are a source of shared tumor rejection antigens for CTL crosspriming. Nat. Med. 7:297–303 121. Machy P, Serre K, Leserman L. 2000. Class I-restricted presentation of exogenous antigen acquired by Fcgamma receptor-mediated endocytosis is regulated in dendritic cells. Eur. J. Immunol. 30:848–57 122. Rovere P, Sabbadini MG, Vallinoto C, Fascio U, Rescigno M, Crosti M, Ricciardi-Castagnoli P, Balestrieri G, Tincani A, Manfredi AA. 1999. Dendritic cell presentation of antigens from apoptotic cells in a proinflammatory context: role of opsonizing anti-beta2-glycoprotein I antibodies. Arthritis Rheumatol. 42:1412–20 123. Wallace PK, Tsang KY, Goldstein J, Correale P, Jarry TM, Schlom J, Guyre PM, Ernstoff MS, Fanger MW. 2001. Exogenous antigen targeted to FcgammaRI on myeloid cells is presented in association

124.

125.

126.

127.

128.

129.

655

with MHC class I. J. Immunol. Methods 248:183–94 Arnold D, Wahl C, Faath S, Rammensee HG, Schild H. 1997. Influences of transporter associated with antigen processing (TAP) on the repertoire of peptides associated with the endoplasmic reticulumresident stress protein gp96. J. Exp. Med. 186:461–66 Binder RJ, Blachere NE, Srivastava PK. 2001. Heat shock protein-chaperoned peptides but not free peptides introduced into the cytosol are presented efficiently by major histocompatibility complex I molecules. J. Biol. Chem. 276:17163–71 Singh-Jasuja H, Toes RE, Spee P, Munz C, Hilf N, Schoenberger SP, RicciardiCastagnoli P, Neefjes J, Rammensee HG, Arnold-Schild D, Schild H. 2000. Cross-presentation of glycoprotein 96associated antigens on major histocompatibility complex class I molecules requires receptor-mediated endocytosis. J. Exp. Med. 191:1965–74 Haicheur N, Bismuth E, Bosset S, Adotevi O, Warnier G, Lacabanne V, Regnault A, Desaymard C, Amigorena S, Ricciardi-Castagnoli P, Goud B, Fridman WH, Johannes L, Tartour E. 2000. The B subunit of Shiga toxin fused to a tumor antigen elicits CTL and targets dendritic cells to allow MHC class I-restricted presentation of peptides derived from exogenous antigens. J. Immunol. 165:3301–8 Guermonprez P, Khelef N, Blouin E, Rieu P, Ricciardi-Castagnoli P, Guiso N, Ladant D, Leclerc C. 2001. The adenylate cyclase toxin of Bordetella pertussis binds to target cells via the alpha(M) beta(2) integrin (CD11b/CD18). J. Exp. Med. 193:1035–44 Yewdell JW, Norbury CC, Bennink JR. 1999. Mechanisms of exogenous antigen presentation by MHC class I molecules in vitro and in vivo: implications for generating CD8+ T cell responses to infectious agents, tumors, transplants, and vaccines. Adv. Immunol. 73:1–77

4 Feb 2002

11:53

Annu. Rev. Immunol. 2002.20:621-667. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

656

AR

AR152-20.tex

AR152-20.SGM

LaTeX2e(2001/05/10)

P1: GJC

GUERMONPREZ ET AL.

130. Gromme M, Uytdehaag FG, Janssen H, Calafat J, van Binnendijk RS, Kenter MJ, Tulp A, Verwoerd D, Neefjes J. 1999. Recycling MHC class I molecules and endosomal peptide loading. Proc. Natl. Acad. Sci. USA 96:10326–31 131. Kleijmeer MJ, Escola JM, UytdeHaag FG, Jakobson E, Griffith JM, Osterhaus AD, Stoorvogel W, Melief CJ, Rabouille C, Geuze HJ. 2001. Antigen loading of MHC class I molecules in the endocytic tract. Traffic 2:124–37 132. Guyre CA, Barreda ME, Swink SL, Fanger MW. 2001. Colocalization of Fc gamma RI-targeted antigen with class I MHC: implications for antigen processing. J. Immunol. 166:2469–78 133. Kovacsovics-Bankowski M, Rock KL. 1995. A phagosome-to-cytosol pathway for exogenous antigens presented on MHC class I molecules. Science 267:243– 46 134. Garbi N, Tan P, Diehl AD, Chambers BJ, Ljunggren HG, Momburg F, Hammerling GJ. 2000. Impaired immune responses and altered peptide repertoire in tapasindeficient mice. Nat. Immunol. 1:234–38 135. Norbury CC, Hewlett LJ, Prescott AR, Shastri N, Watts C. 1995. Class I MHC presentation of exogenous soluble antigen via macropinocytosis in bone marrow macrophages. Immunity 3:783–91 136. Rodriguez A, Regnault A, Kleijmeer M, Ricciardi-Castagnoli P, Amigorena S. 1999. Selective transport of internalized antigens to the cytosol for MHC class I presentation in dendritic cells. Nat. Cell Biol. 1:362–68 137. Merkenschlager M, Power MO, Pircher H, Fisher AG. 1999. Intrathymic deletion of MHC class I-restricted cytotoxic T cell precursors by constitutive crosspresentation of exogenous antigen. Eur. J. Immunol. 29:1477–86 138. Limmer A, Ohl J, Kurts C, Ljunggren HG, Reiss Y, Groettrup M, Momburg F, Arnold B, Knolle PA. 2000. Efficient presentation of exogenous antigen by liver

139.

140.

141.

142.

143.

144.

145.

146.

147.

endothelial cells to CD8+ T cells results in antigen-specific T-cell tolerance. Nat. Med. 6:1348–54 Huang AY, Bruce AT, Pardoll DM, Levitsky HI. 1996. In vivo cross-priming of MHC class I-restricted antigens requires the TAP transporter. Immunity 4:349–55 Sigal LJ, Crotty S, Andino R, Rock KL. 1999. Cytotoxic T-cell immunity to virusinfected non-haematopoietic cells requires presentation of exogenous antigen. Nature 398:77–80 Cresswell P. 1996. Invariant chain structure and MHC class II function. Cell 84:505–7 Villadangos JA, Bryant RA, Deussing J, Driessen C, Lennon-Dumenil AM, Riese RJ, Roth W, Saftig P, Shi GP, Chapman HA, Peters C, Ploegh HL. 1999. Proteases involved in MHC class II antigen presentation. Immunol. Rev. 172:109–20 Kropshofer H, Hammerling GJ, Vogt AB. 1999. The impact of the non-classical MHC proteins HLA-DM and HLA-DO on loading of MHC class II molecules. Immunol. Rev. 172:267–78 Watts C. 2001. Antigen processing in the endocytic compartment. Curr. Opin. Immunol. 13:26–31 Inaba K, Turley S, Iyoda T, Yamaide F, Shimoyama S, Reis e Sousa C, Germain RN, Mellman I, Steinman RM. 2000. The formation of immunogenic major histocompatibility complex class II-peptide ligands in lysosomal compartments of dendritic cells is regulated by inflammatory stimuli. J. Exp. Med. 191:927– 36 Fiebiger E, Meraner P, Weber E, Fang IF, Stingl G, Ploegh H, Maurer D. 2001. Cytokines regulate proteolysis in major histocompatibility complex class II-dependent antigen presentation by dendritic cells. J. Exp. Med. 193:881–92 Bevec T, Stoka V, Pungercic G, Dolenc I, Turk V. 1996. Major histocompatibility complex class II-associated p41 invariant chain fragment is a strong inhibitor

4 Feb 2002

11:53

AR

AR152-20.tex

AR152-20.SGM

LaTeX2e(2001/05/10)

P1: GJC

ANTIGEN PRESENTATION BY DENDRITIC CELLS

148.

Annu. Rev. Immunol. 2002.20:621-667. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

149.

150.

151.

152.

153.

154.

155.

of lysosomal cathepsin L. J. Exp. Med. 183:1331–38 Pierre P, Shachar I, Matza D, Gatti E, Flavell RA, Mellman I. 2000. Invariant chain controls H2-M proteolysis in mouse splenocytes and dendritic cells. J. Exp. Med. 191:1057–62 Pierre P, Turley SJ, Gatti E, Hull M, Meltzer J, Mirza A, Inaba K, Steinman RM, Mellman I. 1997. Developmental regulation of MHC class II transport in mouse dendritic cells. Nature 388:787–92 Pierre P, Mellman I. 1998. Developmental regulation of invariant chain proteolysis controls MHC class II trafficking in mouse dendritic cells. Cell 93:1135– 45 Nakagawa TY, Brissette WH, Lira PD, Griffiths RJ, Petrushova N, Stock J, McNeish JD, Eastman SE, Howard ED, Clarke SR, Rosloniec EF, Elliott EA, Rudensky AY. 1999. Impaired invariant chain degradation and antigen presentation and diminished collagen-induced arthritis in cathepsin S null mice. Immunity 10:207–17 Shi GP, Villadangos JA, Dranoff G, Small C, Gu L, Haley KJ, Riese R, Ploegh HL, Chapman HA. 1999. Cathepsin S required for normal MHC class II peptide loading and germinal center development. Immunity 10:197–206 Driessen C, Bryant RA, Lennon-Dumenil AM, Villadangos JA, Bryant PW, Shi GP, Chapman HA, Ploegh HL. 1999. Cathepsin S controls the trafficking and maturation of MHC class II molecules in dendritic cells. J. Cell Biol. 147:775–90 Rovere P, Zimmermann VS, Forquet F, Demandolx D, Trucy J, Ricciardi-Castagnoli P, Davoust J. 1998. Dendritic cell maturation and antigen presentation in the absence of invariant chain. Proc. Natl. Acad. Sci. USA 95:1067–72 Villadangos JA, Cardoso M, Steptoe RJ, van Berkel D, Pooley J, Carbone FR, Shortman K. 2001. MHC class II expression is regulated in dendritic cells inde-

156.

157.

158.

159.

160.

161.

162.

163.

164.

657

pendently of invariant chain degradation. Immunity 14:739–49 Santambrogio L, Sato AK, Carven GJ, Belyanskaya SL, Strominger JL, Stern LJ. 1999. Extracellular antigen processing and presentation by immature dendritic cells. Proc. Natl. Acad. Sci. USA 96:15056–61 Askew D, Chu RS, Krieg AM, Harding CV. 2000. CpG DNA induces maturation of dendritic cells with distinct effects on nascent and recycling MHCII antigen-processing mechanisms. J. Immunol. 165:6889–95 Turley SJ, Inaba K, Garrett WS, Ebersold M, Unternaehrer J, Steinman RM, Mellman I. 2000. Transport of peptide-MHC class II complexes in developing dendritic cells. Science 288:522–27 Baron C, Raposo G, Scholl SM, Bausinger H, Tenza D, Bohbot A, Pouillart P, Goud B, Hanau D, Salamero J. 2001. Modulation of MHC class II transport and lysosome distribution by macrophage-colony stimulating factor in human dendritic cells derived from monocytes. J. Cell Sci. 114:999–1010 Porcelli SA, Modlin RL. 1999. The CD1 system: antigen-presenting molecules for T cell recognition of lipids and glycolipids. Annu. Rev. Immunol. 17:297–329 Matsuda JL, Kronenberg M. 2001. Presentation of self and microbial lipids by CD1 molecules. Curr. Opin. Immunol. 13:19–25 Jayawardena-Wolf J, Bendelac A. 2001. CD1 and lipid antigens: intracellular pathways for antigen presentation. Curr. Opin. Immunol. 13:109–13 Sugita M, Grant EP, van Donselaar E, Hsu VW, Rogers RA, Peters PJ, Brenner MB. 1999. Separate pathways for antigen presentation by CD1 molecules. Immunity 11:743–52 Salamero J, Bausinger H, Mommaas AM, Lipsker D, Proamer F, Cazenave JP, Goud B, de la Salle H, Hanau D. 2001. CD1a molecules traffic through the

4 Feb 2002

11:53

658

165.

Annu. Rev. Immunol. 2002.20:621-667. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

166.

167.

168.

169.

170.

171.

172.

AR

AR152-20.tex

AR152-20.SGM

LaTeX2e(2001/05/10)

P1: GJC

GUERMONPREZ ET AL. early recycling endosomal pathway in human Langerhans cells. J. Invest. Dermatol. 116:401–8 Sugita M, van Der Wel N, Rogers RA, Peters PJ, Brenner MB. 2000. CD1c molecules broadly survey the endocytic system. Proc. Natl. Acad. Sci. USA 97: 8445–50 Schaible UE, Hagens K, Fischer K, Collins HL, Kaufmann SH. 2000. Intersection of group I CD1 molecules and mycobacteria in different intracellular compartments of dendritic cells. J. Immunol. 164:4843–52 Shamshiev A, Donda A, Prigozy TI, Mori L, Chigorno V, Benedict CA, Kappos L, Sonnino S, Kronenberg M, De Libero G. 2000. The alphabeta T cell response to self-glycolipids shows a novel mechanism of CD1b loading and a requirement for complex oligosaccharides. Immunity 13:255–64 Briken V, Jackman RM, Watts GF, Rogers RA, Porcelli SA. 2000. Human CD1b and CD1c isoforms survey different intracellular compartments for the presentation of microbial lipid antigens. J. Exp. Med. 192:281–88 Park SH, Weiss A, Benlagha K, Kyin T, Teyton L, Bendelac A. 2001. The mouse CD1d-restricted repertoire is dominated by a few autoreactive T cell receptor families. J. Exp. Med. 193:893–904 Gumperz JE, Roy C, Makowska A, Lum D, Sugita M, Podrebarac T, Koezuka Y, Porcelli SA, Cardell S, Brenner MB, Behar SM. 2000. Murine CD1d-restricted T cell recognition of cellular lipids. Immunity 12:211–21 Chiu YH, Jayawardena J, Weiss A, Lee D, Park SH, Dautry-Varsat A, Bendelac A. 1999. Distinct subsets of CD1drestricted T cells recognize self-antigens loaded in different cellular compartments. J. Exp. Med. 189:103–10 Tangri S, Holcombe HR, Castano AR, Miller JE, Teitell M, Huse WE, Peterson PA, Kronenberg M. 1996. Antigen-

173.

174.

175.

176.

177.

178.

presenting function of the mouse CD1 molecule. Ann. N.Y. Acad. Sci. 778:288– 96 Yang OO, Racke FK, Nguyen PT, Gausling R, Severino ME, Horton HF, Byrne MC, Strominger JL, Wilson SB. 2000. CD1d on myeloid dendritic cells stimulates cytokine secretion from and cytolytic activity of V alpha 24J alpha Q T cells: a feedback mechanism for immune regulation. J. Immunol. 165:3756–62 Kitamura H, Iwakabe K, Yahata T, Nishimura S, Ohta A, Ohmi Y, Sato M, Takeda K, Okumura K, Van Kaer L, Kawano T, Taniguchi M, Nishimura T. 1999. The natural killer T (NKT) cell ligand alphagalactosylceramide demonstrates its immunopotentiating effect by inducing interleukin (IL)-12 production by dendritic cells and IL-12 receptor expression on NKT cells. J. Exp. Med. 189:1121–28 Nishimura T, Kitamura H, Iwakabe K, Yahata T, Ohta A, Sato M, Takeda K, Okumura K, Van Kaer L, Kawano T, Taniguchi M, Nakui M, Sekimoto M, Koda T. 2000. The interface between innate and acquired immunity: glycolipid antigen presentation by CD1d-expressing dendritic cells to NKT cells induces the differentiation of antigen-specific cytotoxic T lymphocytes. Int. Immunol. 12: 987–94 Ikarashi Y, Mikami R, Bendelac A, Terme M, Chaput N, Tursz T, Angevin E, Lemonnier FA, Wakasugi H, Zitvogel L. 2001. Dendritic cell maturation overrules H-2D-mediated natural killer (NKT) cell inhibition: critical role for B7 in CD1ddependent NKT cell IFNγ production. J. Exp. Med. 194:1179–86 Angenieux C, Salamero J, Fricker D, Cazenave JP, Goud B, Hanau D, de La Salle H. 2000. Characterization of CD1e, a third type of CD1 molecule expressed in dendritic cells. J. Biol. Chem. 275:37757– 64 Inaba K, Metlay JP, Crowley MT, Steinman RM. 1990. Dendritic cells pulsed

4 Feb 2002

11:53

AR

AR152-20.tex

AR152-20.SGM

LaTeX2e(2001/05/10)

P1: GJC

ANTIGEN PRESENTATION BY DENDRITIC CELLS

Annu. Rev. Immunol. 2002.20:621-667. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

179.

180.

181.

182.

183.

184.

185.

with protein antigens in vitro can prime antigen-specific, MHC-restricted T cells in situ [published erratum appears in J. Exp. Med. [1990 Oct 1; 172(4):1275]. J. Exp. Med. 172:631–40 Mayordomo JI, Zorina T, Storkus WJ, Zitvogel L, Celluzzi C, Falo LD, Melief CJ, Ildstad ST, Kast WM, Deleo AB, et al. 1995. Bone marrow-derived dendritic cells pulsed with synthetic tumour peptides elicit protective and therapeutic antitumour immunity. Nat. Med. 1:1297– 302 Moll H, Fuchs H, Blank C, Rollinghoff M. 1993. Langerhans cells transport Leishmania major from the infected skin to the draining lymph node for presentation to antigen-specific T cells. Eur. J. Immunol. 23:1595–601 Usherwood EJ, Hogg TL, Woodland DL. 1999. Enumeration of antigen-presenting cells in mice infected with Sendai virus. J. Immunol. 162:3350–55 Bujdoso R, Hopkins J, Dutia BM, Young P, McConnell I. 1989. Characterization of sheep afferent lymph dendritic cells and their role in antigen carriage. J. Exp. Med. 170:1285–301 Guery JC, Adorini L. 1995. Dendritic cells are the most efficient in presenting endogenous naturally processed selfepitopes to class II-restricted T cells. J. Immunol. 154:536–44 Zhong G, Sousa CR, Germain RN. 1997. Antigen-unspecific B cells and lymphoid dendritic cells both show extensive surface expression of processed antigenmajor histocompatibility complex class II complexes after soluble protein exposure in vivo or in vitro. J. Exp. Med. 186:673– 82 Reis e Sousa C, Germain RN. 1999. Analysis of adjuvant function by direct visualization of antigen presentation in vivo: endotoxin promotes accumulation of antigen-bearing dendritic cells in the T cell areas of lymphoid tissue. J. Immunol. 162:6552–61

659

186. Casares S, Inaba K, Brumeanu TD, Steinman RM, Bona CA. 1997. Antigen presentation by dendritic cells after immunization with DNA encoding a major histocompatibility complex class II-restricted viral epitope. J. Exp. Med. 186:1481–86 187. Porgador A, Irvine KR, Iwasaki A, Barber BH, Restifo NP, Germain RN. 1998. Predominant role for directly transfected dendritic cells in antigen presentation to CD8+ T cells after gene gun immunization. J. Exp. Med. 188:1075–82 188. Ingulli E, Mondino A, Khoruts A, Jenkins MK. 1997. In vivo detection of dendritic cell antigen presentation to CD4(+) T cells. J. Exp. Med. 185:2133–41 189. Ridge JP, Fuchs EJ, Matzinger P. 1996. Neonatal tolerance revisited: turning on newborn T cells with dendritic cells. Science 271:1723–26 190. Pulendran B, Smith JL, Jenkins M, Schoenborn M, Maraskovsky E, Maliszewski CR. 1998. Prevention of peripheral tolerance by a dendritic cell growth factor: flt3 ligand as an adjuvant. J. Exp. Med. 188:2075–82 191. Steptoe RJ, Fu F, Li W, Drakes ML, Lu L, Demetris AJ, Qian S, McKenna HJ, Thomson AW. 1997. Augmentation of dendritic cells in murine organ donors by Flt3 ligand alters the balance between transplant tolerance and immunity. J. Immunol. 159:5483–91 192. Mayordomo JI, Loftus DJ, Sakamoto H, De Cesare CM, Appasamy PM, Lotze MT, Storkus WJ, Appella E, DeLeo AB. 1996. Therapy of murine tumors with p53 wild-type and mutant sequence peptidebased vaccines. J. Exp. Med. 183:1357– 65 193. Ludewig B, Odermatt B, Landmann S, Hengartner H, Zinkernagel RM. 1998. Dendritic cells induce autoimmune diabetes and maintain disease via de novo formation of local lymphoid tissue. J. Exp. Med. 188:1493–501 194. Schreurs MW, Eggert AA, de Boer AJ,

4 Feb 2002

11:53

660

Annu. Rev. Immunol. 2002.20:621-667. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

195.

196.

197.

198.

199.

200.

201.

AR

AR152-20.tex

AR152-20.SGM

LaTeX2e(2001/05/10)

P1: GJC

GUERMONPREZ ET AL. Vissers JL, van Hall T, Offringa R, Figdor CG, Adema GJ. 2000. Dendritic cells break tolerance and induce protective immunity against a melanocyte differentiation antigen in an autologous melanoma model. Cancer Res. 60:6995– 7001 Ludewig B, Ochsenbein AF, Odermatt B, Paulin D, Hengartner H, Zinkernagel RM. 2000. Immunotherapy with dendritic cells directed against tumor antigens shared with normal host cells results in severe autoimmune disease. J. Exp. Med. 191:795–804 Brocker T, Riedinger M, Karjalainen K. 1997. Targeted expression of major histocompatibility complex (MHC) class II molecules demonstrates that dendritic cells can induce negative but not positive selection of thymocytes in vivo. J. Exp. Med. 185:541–50 Kyewski BA, Fathman CG, Rouse RV. 1986. Intrathymic presentation of circulating non-MHC antigens by medullary dendritic cells. An antigen-dependent microenvironment for T cell differentiation. J. Exp. Med. 163:231–46 Crowley M, Inaba K, Steinman RM. 1990. Dendritic cells are the principal cells in mouse spleen bearing immunogenic fragments of foreign proteins. J. Exp. Med. 172:383–86 Guery JC, Ria F, Galbiati F, Smiroldo S, Adorini L. 1997. The mode of protein antigen administration determines preferential presentation of peptide-class II complexes by lymph node dendritic or B cells. Int. Immunol. 9:9–15 Inaba K, Pack M, Inaba M, Sakuta H, Isdell F, Steinman RM. 1997. High levels of a major histocompatibility complex II-self peptide complex on dendritic cells from the T cell areas of lymph nodes. J. Exp. Med. 186:665– 72 Finkelman FD, Lees A, Birnbaum R, Gause WC, Morris SC. 1996. Dendritic cells can present antigen in vivo in a

201a.

202.

203.

204.

205.

206.

207.

208.

tolerogenic or immunogenic fashion. J. Immunol. 157:1406–14 Hawiger D, Inaba K, Dorsett Y, Guo M, Mahnke K, Rivera M, Ravetch JV, Steinman RM, Nussenzweig MC. 2001. Dendritic cells induce peripheral T cell unresponsiveness under steady state conditions in vivo. J. Exp. Med. 194:769–79 Thomson AW, Lu L. 1999. Are dendritic cells the key to liver transplant tolerance? Immunol. Today 20:27–32 Huang AY, Golumbek P, Ahmadzadeh M, Jaffee E, Pardoll D, Levitsky H. 1994. Role of bone marrow-derived cells in presenting MHC class I-restricted tumor antigens. Science 264:961–65 Lenz LL, Butz EA, Bevan MJ. 2000. Requirements for bone marrow-derived antigen-presenting cells in priming cytotoxic T cell responses to intracellular pathogens. J. Exp. Med. 192:1135–42 Sigal LJ, Rock KL. 2000. Bone marrowderived antigen-presenting cells are required for the generation of cytotoxic T lymphocyte responses to viruses and use transporter associated with antigen presentation (TAP)-dependent and -independent pathways of antigen presentation. J. Exp. Med. 192:1143–50 Miller JF, Kurts C, Allison J, Kosaka H, Carbone F, Heath WR. 1998. Induction of peripheral CD8+ T-cell tolerance by cross-presentation of self antigens. Immunol. Rev. 165:267–77 Kurts C, Cannarile M, Klebba I, Brocker T. 2001. Dendritic cells are sufficient to cross-present self-antigens to CD8 T cells in vivo. J. Immunol. 166:1439–42 Chiodoni C, Paglia P, Stoppacciaro A, Rodolfo M, Parenza M, Colombo MP. 1999. Dendritic cells infiltrating tumors cotransduced with granulocyte/ macrophage colony-stimulating factor (GM-CSF) and CD40 ligand genes take up and present endogenous tumorassociated antigens, and prime naive mice for a cytotoxic T lymphocyte response. J. Exp. Med. 190:125–33

4 Feb 2002

11:53

AR

AR152-20.tex

AR152-20.SGM

LaTeX2e(2001/05/10)

P1: GJC

Annu. Rev. Immunol. 2002.20:621-667. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

ANTIGEN PRESENTATION BY DENDRITIC CELLS 209. Li M, Davey GM, Sutherland RM, Kurts C, Lew AM, Hirst C, Carbone FR, Heath WR. 2001. Cell-associated ovalbumin is cross-presented much more efficiently than soluble ovalbumin in vivo. J. Immunol. 166:6099–103 210. Banchereau J, Pulendran B, Steinman R, Palucka K. 2000. Will the making of plasmacytoid dendritic cells in vitro help unravel their mysteries? J. Exp. Med. 192:F39–44 211. Shortman K, Wu L. 2001. Parentage and heritage of dendritic cells. Blood 97:3325 212. Ruedl C, Bachmann MF. 1999. CTL priming by CD8(+) and CD8(−) dendritic cells in vivo. Eur. J. Immunol. 29: 3762–67 213. Pooley JL, Heath WR, Shortman K. 2001. Cutting edge: intravenous soluble antigen is presented to cd4 t cells by CD8(−) dendritic cells, but crosspresented to CD8 T cells by CD8(+) dendritic cells. J. Immunol. 166:5327–30 214. den Haan JM, Lehar SM, Bevan MJ. 2000. CD8(+) but not CD8(−) dendritic cells cross-prime cytotoxic T cells in vivo. J. Exp. Med. 192:1685–96 215. Maldonado-Lopez R, De Smedt T, Michel P, Godfroid J, Pajak B, Heirman C, Thielemans K, Leo O, Urbain J, Moser M. 1999. CD8alpha+ and CD8alpha−− subclasses of dendritic cells direct the development of distinct T helper cells in vivo. J. Exp. Med. 189: 587–92 216. Pulendran B, Smith JL, Caspary G, Brasel K, Pettit D, Maraskovsky E, Maliszewski CR. 1999. Distinct dendritic cell subsets differentially regulate the class of immune response in vivo. Proc. Natl. Acad. Sci. USA 96:1036–41 217. Suss G, Shortman K. 1996. A subclass of dendritic cells kills CD4 T cells via Fas/Fas-ligand- induced apoptosis. J. Exp. Med. 183:1789–96 218. Smith AL, de St Groth BF. 1999. Antigen-pulsed CD8alpha+ dendritic

219.

220.

221.

222.

223.

224.

225.

661

cells generate an immune response after subcutaneous injection without homing to the draining lymph node. J. Exp. Med. 189:593–98 Liu YJ, Kanzler H, Soumelis V, Gilliet M. 2001. Dendritic cell lineage, plasticity and cross-regulation. Nat. Immunol. 2:585–89 Grouard G, Rissoan MC, Filgueira L, Durand I, Banchereau J, Liu YJ. 1997. The enigmatic plasmacytoid T cells develop into dendritic cells with interleukin (IL)-3 and CD40-ligand. J. Exp. Med. 185:1101–11 Rissoan MC, Soumelis V, Kadowaki N, Grouard G, Briere F, de Waal Malefyt R, Liu YJ. 1999. Reciprocal control of T helper cell and dendritic cell differentiation. Science 283:1183–86 Freeman GJ, Long AJ, Iwai Y, Bourque K, Chernova T, Nishimura H, Fitz LJ, Malenkovich N, Okazaki T, Byrne MC, Horton HF, Fouser L, Carter L, Ling V, Bowman MR, Carreno BM, Collins M, Wood CR, Honjo T. 2000. Engagement of the PD-1 immunoinhibitory receptor by a novel B7 family member leads to negative regulation of lymphocyte activation. J. Exp. Med. 192:1027– 34 Dong H, Zhu G, Tamada K, Chen L. 1999. B7-H1, a third member of the B7 family, co-stimulates T-cell proliferation and interleukin-10 secretion. Nat. Med. 5:1365–69 Tseng SY, Otsuji M, Gorski K, Huang X, Slansky JE, Pai SI, Shalabi A, Shin T, Pardoll DM, Tsuchiya H. 2001. B7-DC, a new dendritic cell molecule with potent costimulatory properties for T cells. J. Exp. Med. 193:839–46 Shuford WW, Klussman K, Tritchler DD, Loo DT, Chalupny J, Siadak AW, Brown TJ, Emswiler J, Raecho H, Larsen CP, Pearson TC, Ledbetter JA, Aruffo A, Mittler RS. 1997. 4-1BB costimulatory signals preferentially induce CD8+ T cell proliferation and lead to the

4 Feb 2002

11:53

662

226.

Annu. Rev. Immunol. 2002.20:621-667. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

227.

228.

229.

230.

231.

232.

233.

234.

AR

AR152-20.tex

AR152-20.SGM

LaTeX2e(2001/05/10)

P1: GJC

GUERMONPREZ ET AL. amplification in vivo of cytotoxic T cell responses. J. Exp. Med. 186:47–55 Lane P. 2000. Role of OX40 signals in coordinating CD4 T cell selection, migration, and cytokine differentiation in T helper (Th)1 and Th2 cells. J. Exp. Med. 191:201–6 Ruedl C, Kopf M, Bachmann MF. 1999. CD8(+) T cells mediate CD40-independent maturation of dendritic cells in vivo. J. Exp. Med. 189:1875–84 Bennett SR, Carbone FR, Karamalis F, Miller JF, Heath WR. 1997. Induction of a CD8+ cytotoxic T lymphocyte response by cross-priming requires cognate CD4+ T cell help. J. Exp. Med. 186:65–70 Kurts C, Carbone FR, Barnden M, Blanas E, Allison J, Heath WR, Miller JF. 1997. CD4+ T cell help impairs CD8+ T cell deletion induced by cross-presentation of self-antigens and favors autoimmunity. J. Exp. Med. 186:2057–62 Bennett SR, Carbone FR, Karamalis F, Flavell RA, Miller JF, Heath WR. 1998. Help for cytotoxic-T-cell responses is mediated by CD40 signalling. Nature 393:478–80 Ridge JP, Di Rosa F, Matzinger P. 1998. A conditioned dendritic cell can be a temporal bridge between a CD4+ T-helper and a T-killer cell. Nature 393: 474–78 Schoenberger SP, Toes RE, van der Voort EI, Offringa R, Melief CJ. 1998. T-cell help for cytotoxic T lymphocytes is mediated by CD40-CD40L interactions. Nature 393:480–83 Mackey MF, Gunn JR, Maliszewsky C, Kikutani H, Noelle RJ, Barth RJ Jr. 1998. Dendritic cells require maturation via CD40 to generate protective antitumor immunity. J. Immunol. 161:2094– 98 Diehl L, den Boer AT, Schoenberger SP, van der Voort EI, Schumacher TN, Melief CJ, Offringa R, Toes RE. 1999. CD40 activation in vivo overcomes

235.

236.

237.

238.

239.

240.

241.

242.

peptide-induced peripheral cytotoxic Tlymphocyte tolerance and augments anti-tumor vaccine efficacy. Nat. Med. 5:774–79 Lu Z, Yuan L, Zhou X, Sotomayor E, Levitsky HI, Pardoll DM. 2000. CD40independent pathways of T cell help for priming of CD8(+) cytotoxic T lymphocytes. J. Exp. Med. 191:541–50 Hirao M, Onai N, Hiroishi K, Watkins SC, Matsushima K, Robbins PD, Lotze MT, Tahara H. 2000. CC chemokine receptor-7 on dendritic cells is induced after interaction with apoptotic tumor cells: critical role in migration from the tumor site to draining lymph nodes. Cancer Res. 60:2209–17 Steinman RM, Turley S, Mellman I, Inaba K. 2000. The induction of tolerance by dendritic cells that have captured apoptotic cells. J. Exp. Med. 191:411– 16 Roncarolo MG, Levings MK, Traversari C. 2001. Differentiation of T regulatory cells by immature dendritic cells. J. Exp. Med. 193:F5–9 Jonuleit H, Schmitt E, Steinbrink K, Enk AH. 2001. Dendritic cells as a tool to induce anergic and regulatory T cells. Trends Immunol. 22:394–400 Jonuleit H, Schmitt E, Schuler G, Knop J, Enk AH. 2000. Induction of interleukin 10-producing, nonproliferating CD4(+) T cells with regulatory properties by repetitive stimulation with allogeneic immature human dendritic cells. J. Exp. Med. 192:1213–22 Dhodapkar MV, Steinman RM, Krasovsky J, Munz C, Bhardwaj N. 2001. Antigen-specific inhibition of effector T cell function in humans after injection of immature dendritic cells. J. Exp. Med. 193:233–38 Fu F, Li Y, Qian S, Lu L, Chambers F, Starzl TE, Fung JJ, Thomson AW. 1996. Costimulatory molecule-deficient dendritic cell progenitors (MHC class II+, CD80dim, CD86−−) prolong

4 Feb 2002

11:53

AR

AR152-20.tex

AR152-20.SGM

LaTeX2e(2001/05/10)

P1: GJC

ANTIGEN PRESENTATION BY DENDRITIC CELLS

Annu. Rev. Immunol. 2002.20:621-667. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

243.

244.

245.

246.

247.

248.

249.

250.

251.

cardiac allograft survival in nonimmunosuppressed recipients. Transplantation 62:659–65 Cella M, Scheidegger D, Palmer-Lehmann K, Lane P, Lanzavecchia A, Alber G. 1996. Ligation of CD40 on dendritic cells triggers production of high levels of interleukin-12 and enhances T cell stimulatory capacity: T-T help via APC activation. J. Exp. Med. 184:747–52 Hartmann G, Weiner GJ, Krieg AM. 1999. CpG DNA: a potent signal for growth, activation, and maturation of human dendritic cells. Proc. Natl. Acad. Sci. USA 96:9305–10 Cella M, Jarrossay D, Facchetti F, Alebardi O, Nakajima H, Lanzavecchia A, Colonna M. 1999. Plasmacytoid monocytes migrate to inflamed lymph nodes and produce large amounts of type I interferon. Nat. Med. 5:919–23 Cella M, Facchetti F, Lanzavecchia A, Colonna M. 2000. Plasmacytoid dendritic cells activated by influenza virus and CD40L drive a potent TH1 polarization. Nat. Immunol. 1:305–10 Kadowaki N, Antonenko S, Lau JY, Liu YJ. 2000. Natural interferon alpha/betaproducing cells link innate and adaptive immunity. J. Exp. Med. 192:219–26 Pulendran B, Palucka K, Banchereau J. 2001. Sensing pathogens and tuning immune responses. Science 293:253–56 Iwasaki A, Kelsall BL. 1999. Freshly isolated Peyer’s patch, but not spleen, dendritic cells produce interleukin 10 and induce the differentiation of T helper type 2 cells. J. Exp. Med. 190:229–39 Stumbles PA, Thomas JA, Pimm CL, Lee PT, Venaille TJ, Proksch S, Holt PG. 1998. Resting respiratory tract dendritic cells preferentially stimulate T helper cell type 2 (Th2) responses and require obligatory cytokine signals for induction of Th1 immunity. J. Exp. Med. 188:2019–31 Tanaka H, Demeure CE, Rubio M, Delespesse G, Sarfati M. 2000. Human

252.

253.

254.

255.

256.

257.

258.

259.

663

monocyte-derived dendritic cells induce naive T cell differentiation into T helper cell type 2 (Th2) or Th1/Th2 effectors. Role of stimulator/responder ratio. J. Exp. Med. 192:405–12 Langenkamp A, Messi M, Lanzavecchia A, Sallusto F. 2000. Kinetics of dendritic cell activation: impact on priming of TH1, TH2 and nonpolarized T cells. Nat. Immunol. 1:311–16 Iezzi G, Karjalainen K, Lanzavecchia A. 1998. The duration of antigenic stimulation determines the fate of naive and effector T cells. Immunity 8:89–95 Bevan MJ, Fink PJ. 2001. The CD8 response on autopilot. Nat. Immunol. 2: 381–82 Josien BR, Li HL, Ingulli E, Sarma S, Wong BR, Vologodskaia M, Steinman RM, Choi Y. 2000. TRANCE, a tumor necrosis factor family member, enhances the longevity and adjuvant properties of dendritic cells in vivo. J. Exp. Med. 191: 495–502 De Smedt T, Pajak B, Klaus GG, Noelle RJ, Urbain J, Leo O, Moser M. 1998. Antigen-specific T lymphocytes regulate lipopolysaccharide-induced apoptosis of dendritic cells in vivo. J. Immunol. 161:4476–79 Hermans IF, Ritchie DS, Yang J, Roberts JM, Ronchese F. 2000. CD8+ T cell-dependent elimination of dendritic cells in vivo limits the induction of antitumor immunity. J. Immunol. 164:3095– 101 Ludewig B, Bonilla WV, Dumrese T, Odermatt B, Zinkernagel RM, Hengartner H. 2001. Perforin-independent regulation of dendritic cell homeostasis by CD8(+) T cells in vivo: implications for adaptive immunotherapy. Eur. J. Immunol. 31:1772–79 Ludewig B, Oehen S, Barchiesi F, Schwendener RA, Hengartner H, Zinkernagel RM. 1999. Protective antiviral cytotoxic T cell memory is most efficiently maintained by restimulation via

4 Feb 2002

11:53

664

260.

Annu. Rev. Immunol. 2002.20:621-667. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

261.

262.

263.

264.

265.

266.

AR

AR152-20.tex

AR152-20.SGM

LaTeX2e(2001/05/10)

P1: GJC

GUERMONPREZ ET AL. dendritic cells. J. Immunol. 163:1839– 44 McLellan A, Heldmann M, Terbeck G, Weih F, Linden C, Brocker EB, Leverkus M, Kampgen E. 2000. MHC class II and CD40 play opposing roles in dendritic cell survival. Eur. J. Immunol. 30:2612–19 Anderson DM, Maraskovsky E, Billingsley WL, Dougall WC, Tometsko ME, Roux ER, Teepe MC, DuBose RF, Cosman D, Galibert L. 1997. A homologue of the TNF receptor and its ligand enhance T-cell growth and dendritic-cell function. Nature 390:175–79 Wong BR, Josien R, Lee SY, Sauter B, Li HL, Steinman RM, Choi Y. 1997. TRANCE (tumor necrosis factor [TNF]related activation-induced cytokine), a new TNF family member predominantly expressed in T cells, is a dendritic cellspecific survival factor. J. Exp. Med. 186: 2075–80 Koch F, Heufler C, Kampgen E, Schneeweiss D, Bock G, Schuler G. 1990. Tumor necrosis factor alpha maintains the viability of murine epidermal Langerhans cells in culture, but in contrast to granulocyte/macrophage colonystimulating factor, without inducing their functional maturation. J. Exp. Med. 171:159–71 Wang B, Fujisawa H, Zhuang L, Kondo S, Shivji GM, Kim CS, Mak TW, Sauder DN. 1997. Depressed Langerhans cell migration and reduced contact hypersensitivity response in mice lacking TNF receptor p75. J. Immunol. 159:6148–55 Funk JO, Walczak H, Voigtlander C, Berchtold S, Baumeister T, Rauch P, Rossner S, Steinkasserer A, Schuler G, Lutz MB. 2000. Cutting edge: resistance to apoptosis and continuous proliferation of dendritic cells deficient for TNF receptor-1. J. Immunol. 165:4792–96 Ashany D, Savir A, Bhardwaj N, Elkon KB. 1999. Dendritic cells are resistant

267.

268.

269.

270.

271.

272.

to apoptosis through the Fas (CD95/ APO-1) pathway. J. Immunol. 163: 5303–11 Leverkus M, Walczak H, McLellan A, Fries HW, Terbeck G, Brocker EB, Kampgen E. 2000. Maturation of dendritic cells leads to up-regulation of cellular FLICE-inhibitory protein and concomitant down-regulation of death ligand-mediated apoptosis. Blood 96:2628–31 McLellan AD, Terbeck G, Mengling T, Starling GC, Kiener PA, Gold R, Brocker EB, Leverkus M, Kampgen E. 2000. Differential susceptibility to CD95 (Apo-1/Fas) and MHC class II- induced apoptosis during murine dendritic cell development. Cell Death Differ. 7:933– 38 Bladergroen BA, Strik MC, Bovenschen N, van Berkum O, Scheffer GL, Meijer CJ, Hack CE, Kummer JA. 2001. The granzyme B inhibitor, protease inhibitor 9, is mainly expressed by dendritic cells and at immune-privileged sites. J. Immunol. 166:3218–25 Medema JP, Schouurhuis D, Van Tongeren J, De Jong J, Bres S, Toes R, Toebes M, Schumacher T, BladerGroen B, Ossendorp F, Kummer J, Melief C, Offringa R. 2001. Expression of the serpin SPI-6 protects dendritic cells from T lymphocyte-induced apoptosis: differential modulation by T-helper type 1 and type 2 cells. Presented at Keystone Conference, Taos Mackensen A, Drager R, Schlesier M, Mertelsmann R, Lindemann A. 2000. Presence of IgE antibodies to bovine serum albumin in a patient developing anaphylaxis after vaccination with human peptide-pulsed dendritic cells. Cancer Immunol. Immunother. 49:152–56 Morse MA, Nair S, Fernandez-Casal M, Deng Y, St Peter M, Williams R, Hobeika A, Mosca P, Clay T, Cumming RI, Fisher E, Clavien P, Proia AD, Niedzwiecki D, Caron D, Lyerly HK.

4 Feb 2002

11:53

AR

AR152-20.tex

AR152-20.SGM

LaTeX2e(2001/05/10)

P1: GJC

ANTIGEN PRESENTATION BY DENDRITIC CELLS

Annu. Rev. Immunol. 2002.20:621-667. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

273.

274.

275.

276.

277.

278.

279.

2000. Preoperative mobilization of circulating dendritic cells by Flt3 ligand administration to patients with metastatic colon cancer. J. Clin. Oncol. 18:3883–93 Choudhury BA, Liang JC, Thomas EK, Flores-Romo L, Xie QS, Agusala K, Sutaria S, Sinha I, Champlin RE, Claxton DF. 1999. Dendritic cells derived in vitro from acute myelogenous leukemia cells stimulate autologous, antileukemic T-cell responses. Blood 93:780–86 Charbonnier A, Gaugler B, Sainty D, Lafage-Pochitaloff M, Olive D. 1999. Human acute myeloblastic leukemia cells differentiate in vitro into mature dendritic cells and induce the differentiation of cytotoxic T cells against autologous leukemias. Eur. J. Immunol. 29:2567–78 Eibl B, Ebner S, Duba C, Bock G, Romani N, Erdel M, Gachter A, Niederwieser D, Schuler G. 1997. Dendritic cells generated from blood precursors of chronic myelogenous leukemia patients carry the Philadelphia translocation and can induce a CML-specific primary cytotoxic T-cell response. Genes Chromosomes Cancer 20:215–23 Boczkowski D, Nair SK, Snyder D, Gilboa E. 1996. Dendritic cells pulsed with RNA are potent antigen-presenting cells in vitro and in vivo. J. Exp. Med. 184:465–72 Nair SK, Heiser A, Boczkowski D, Majumdar A, Naoe M, Lebkowski JS, Vieweg J, Gilboa E. 2000. Induction of cytotoxic T cell responses and tumor immunity against unrelated tumors using telomerase reverse transcriptase RNA transfected dendritic cells. Nat. Med. 6:1011–17 Jenne L, Schuler G, Steinkasserer A. 2001. Viral vectors for dendritic cellbased immunotherapy. Trends Immunol. 22:102–7 Irvine AS, Trinder PK, Laughton DL, Ketteringham H, McDermott RH, Reid

280.

281.

282.

283.

284.

285.

286.

287.

665

SC, Haines AM, Amir A, Husain R, Doshi R, Young LS, Mountain A. 2000. Efficient nonviral transfection of dendritic cells and their use for in vivo immunization. Nat. Biotechnol. 18:1273– 78 Gong J, Avigan D, Chen D, Wu Z, Koido S, Kashiwaba M, Kufe D. 2000. Activation of antitumor cytotoxic T lymphocytes by fusions of human dendritic cells and breast carcinoma cells. Proc. Natl. Acad. Sci. USA 97:2715–18 Shu S, Cohen P. 2001. Tumor-dendritic cell fusion technology and immunotherapy strategies. J. Immunother. 24:99– 100 Celluzzi CM, Mayordomo JI, Storkus WJ, Lotze MT, Falo LD Jr. 1996. Peptide-pulsed dendritic cells induce antigen-specific CTL-mediated protective tumor immunity. J. Exp. Med. 183:283– 87 Porgador A, Snyder D, Gilboa E. 1996. Induction of antitumor immunity using bone marrow-generated dendritic cells. J. Immunol. 156:2918–26 Zitvogel L, Mayordomo JI, Tjandrawan T, DeLeo AB, Clarke MR, Lotze MT, Storkus WJ. 1996. Therapy of murine tumors with tumor peptide-pulsed dendritic cells: dependence on T cells, B7 costimulation, and T helper cell 1-associated cytokines. J. Exp. Med. 183:87–97 Alters SE, Gadea JR, Philip R. 1997. Immunotherapy of cancer. Generation of CEA specific CTL using CEA peptide pulsed dendritic cells. Adv. Exp. Med. Biol. 417:519–24 Tjandrawan T, Martin DM, Maeurer MJ, Castelli C, Lotze MT, Storkus WJ. 1998. Autologous human dendriphages pulsed with synthetic or natural tumor peptides elicit tumor-specific CTLs in vitro. J. Immunother. 21:149–57 Bakker AB, Marland G, de Boer AJ, Huijbens RJ, Danen EH, Adema GJ, Figdor CG. 1995. Generation of antimelanoma cytotoxic T lymphocytes

4 Feb 2002

11:53

666

Annu. Rev. Immunol. 2002.20:621-667. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

288.

289.

290.

291.

292.

293.

294.

AR

AR152-20.tex

AR152-20.SGM

LaTeX2e(2001/05/10)

P1: GJC

GUERMONPREZ ET AL. from healthy donors after presentation of melanoma-associated antigen-derived epitopes by dendritic cells in vitro. Cancer Res. 55:5330–34 Subklewe M, Chahroudi A, Bickham K, Larsson M, Kurilla MG, Bhardwaj N, Steinman RM. 1999. Presentation of Epstein-Barr virus latency antigens to CD8(+), interferon-gamma-secreting, T lymphocytes. Eur. J. Immunol. 29:3995– 4001 Amoscato AA, Prenovitz DA, Lotze MT. 1998. Rapid extracellular degradation of synthetic class I peptides by human dendritic cells. J. Immunol. 161:4023–32 Ludewig B, McCoy K, Pericin M, Ochsenbein AF, Dumrese T, Odermatt B, Toes RE, Melief CJ, Hengartner H, Zinkernagel RM. 2001. Rapid peptide turnover and inefficient presentation of exogenous antigen critically limit the activation of self-reactive CTL by dendritic cells. J. Immunol. 166:3678– 87 Jager E, Ringhoffer M, Karbach J, Arand M, Oesch F, Knuth A. 1996. Inverse relationship of melanocyte differentiation antigen expression in melanoma tissues and CD8+ cytotoxicT-cell responses: evidence for immunoselection of antigen-loss variants in vivo. Int. J. Cancer 66:470–76 Riker A, Cormier J, Panelli M, Kammula U, Wang E, Abati A, Fetsch P, Lee KH, Steinberg S, Rosenberg S, Marincola F. 1999. Immune selection after antigen-specific immunotherapy of melanoma. Surgery 126:112–20 Albert ML, Darnell JC, Bender A, Francisco LM, Bhardwaj N, Darnell RB. 1998. Tumor-specific killer cells in paraneoplastic cerebellar degeneration. Nat. Med. 4:1321–24 Shen H, Miller JF, Fan X, Kolwyck D, Ahmed R, Harty JT. 1998. Compartmentalization of bacterial antigens: differential effects on priming of CD8 T cells

295.

296.

297.

298.

299.

300.

301.

and protective immunity. Cell 92:535– 45 Blake N, Lee S, Redchenko I, Thomas W, Steven N, Leese A, SteigerwaldMullen P, Kurilla MG, Frappier L, Rickinson A. 1997. Human CD8+ T cell responses to EBV EBNA1: HLA class I presentation of the (Gly-Ala)-containing protein requires exogenous processing. Immunity 7:791–802 Dhodapkar MV, Steinman RM, Sapp M, Desai H, Fossella C, Krasovsky J, Donahoe SM, Dunbar PR, Cerundolo V, Nixon DF, Bhardwaj N. 1999. Rapid generation of broad T-cell immunity in humans after a single injection of mature dendritic cells. J. Clin. Invest. 104:173– 80 Zitvogel L, Angevin E, Tursz T. 2000. Dendritic cell-based immunotherapy of cancer. Ann. Oncol. 11 (Suppl.) 3:199– 205 Nestle FO, Alijagic S, Gilliet M, Sun Y, Grabbe S, Dummer R, Burg G, Schadendorf D. 1998. Vaccination of melanoma patients with peptide- or tumor lysate-pulsed dendritic cells. Nat. Med. 4:328–32 Murphy GP, Tjoa BA, Simmons SJ, Ragde H, Rogers M, Elgamal A, Kenny GM, Troychak MJ, Salgaller ML, Boynton AL. 1999. Phase II prostate cancer vaccine trial: report of a study involving 37 patients with disease recurrence following primary treatment. Prostate 39:54–59 Panelli MC, Wunderlich J, Jeffries J, Wang E, Mixon A, Rosenberg SA, Marincola FM. 2000. Phase 1 study in patients with metastatic melanoma of immunization with dendritic cells presenting epitopes derived from the melanoma-associated antigens MART-1 and gp100. J. Immunother. 23:487–98 Reichardt VL, Okada CY, Liso A, Benike CJ, Stockerl-Goldstein KE, Engleman EG, Blume KG, Levy R. 1999. Idiotype vaccination using dendritic

4 Feb 2002

11:53

AR

AR152-20.tex

AR152-20.SGM

LaTeX2e(2001/05/10)

P1: GJC

ANTIGEN PRESENTATION BY DENDRITIC CELLS

Annu. Rev. Immunol. 2002.20:621-667. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

302.

303.

304.

305.

306.

cells after autologous peripheral blood stem cell transplantation for multiple myeloma—a feasibility study. Blood 93:2411–19 Small EJ, Fratesi P, Reese DM, Strang G, Laus R, Peshwa MV, Valone FH. 2000. Immunotherapy of hormonerefractory prostate cancer with antigenloaded dendritic cells. J. Clin. Oncol. 18: 3894–903 Fong L, Hou Y, Rivas A, Benike C, Yuen A, Fisher GA, Davis MM, Engleman EG. 2001. Altered peptide ligand vaccination with Flt3 ligand expanded dendritic cells for tumor immunotherapy. Proc. Natl. Acad. Sci. USA 98:8809– 14 Thurner B, Haendle I, Roder C, Dieckmann D, Keikavoussi P, Jonuleit H, Bender A, Maczek C, Schreiner D, von den Driesch P, Brocker EB, Steinman RM, Enk A, Kampgen E, Schuler G. 1999. Vaccination with mage-3A1 peptide-pulsed mature, monocyte-derived dendritic cells expands specific cytotoxic T cells and induces regression of some metastases in advanced stage IV melanoma. J. Exp. Med. 190:1669–78 Holtl L, Rieser C, Papesh C, Ramoner R, Herold M, Klocker H, Radmayr C, Stenzl A, Bartsch G, Thurnher M. 1999. Cellular and humoral immune responses in patients with metastatic renal cell carcinoma after vaccination with antigenpulsed dendritic cells. J. Urol. 161:777– 82 Mackensen A, Herbst B, Chen JL, Kohler G, Noppen C, Herr W, Spagnoli GC, Cerundolo V, Lindemann A. 2000. Phase I study in melanoma patients of

307.

308.

309.

310.

311.

667

a vaccine with peptide-pulsed dendritic cells generated in vitro from CD34(+) hematopoietic progenitor cells. Int. J. Cancer 86:385–92 Kugler A, Stuhler G, Walden P, Zoller G, Zobywalski A, Brossart P, Trefzer U, Ullrich S, Muller CA, Becker V, Gross AJ, Hemmerlein B, Kanz L, Muller GA, Ringert RH. 2000. Regression of human metastatic renal cell carcinoma after vaccination with tumor cell-dendritic cell hybrids. Nat. Med. 6:332–36 Kundu SK, Engleman E, Benike C, Shapero MH, Dupuis M, van Schooten WC, Eibl M, Merigan TC. 1998. A pilot clinical trial of HIV antigen-pulsed allogeneic and autologous dendritic cell therapy in HIV-infected patients. AIDS Res. Hum. Retroviruses 14:551–60 Zitvogel L, Regnault A, Lozier A, Wolfers J, Flament C, Tenza D, RicciardiCastagnoli P, Raposo G, Amigorena S. 1998. Eradication of established murine tumors using a novel cell-free vaccine: dendritic cell-derived exosomes. Nat. Med. 4:594–600 Thery C, Regnault A, Garin J, Wolfers J, Zitvogel L, Ricciardi-Castagnoli P, Raposo G, Amigorena S. 1999. Molecular characterization of dendritic cell-derived exosomes. Selective accumulation of the heat shock protein hsc73. J. Cell Biol. 147:599–610 Thery C, Boussac M, Veron P, Ricciardi-Castagnoli P, Raposo G, Garin J, Amigorena S. 2001. Proteomic analysis of dendritic cell-derived exosomes: a secreted subcellular compartment distinct from apoptotic vesicles. J. Immunol. 166:7309–18

P1: FRK

February 1, 2002

12:47

Annual Reviews

AR152-FM

Annual Review of Immunology Volume 20, 2002

Annu. Rev. Immunol. 2002.20:621-667. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

CONTENTS FRONTISPIECE, Charles A. Janeway, Jr. A TRIP THROUGH MY LIFE WITH AN IMMUNOLOGICAL THEME, Charles A. Janeway, Jr.

xiv 1

THE B7 FAMILY OF LIGANDS AND ITS RECEPTORS: NEW PATHWAYS FOR COSTIMULATION AND INHIBITION OF IMMUNE RESPONSES, Beatriz M. Carreno and Mary Collins

29

MAP KINASES IN THE IMMUNE RESPONSE, Chen Dong, Roger J. Davis, and Richard A. Flavell

PROSPECTS FOR VACCINE PROTECTION AGAINST HIV-1 INFECTION AND AIDS, Norman L. Letvin, Dan H. Barouch, and David C. Montefiori T CELL RESPONSE IN EXPERIMENTAL AUTOIMMUNE ENCEPHALOMYELITIS (EAE): ROLE OF SELF AND CROSS-REACTIVE ANTIGENS IN SHAPING, TUNING, AND REGULATING THE AUTOPATHOGENIC T CELL REPERTOIRE, Vijay K. Kuchroo, Ana C. Anderson, Hanspeter Waldner, Markus Munder, Estelle Bettelli, and Lindsay B. Nicholson

55 73

101

NEUROENDOCRINE REGULATION OF IMMUNITY, Jeanette I. Webster, Leonardo Tonelli, and Esther M. Sternberg

125

MOLECULAR MECHANISM OF CLASS SWITCH RECOMBINATION: LINKAGE WITH SOMATIC HYPERMUTATION, Tasuku Honjo, Kazuo Kinoshita, and Masamichi Muramatsu

165

INNATE IMMUNE RECOGNITION, Charles A. Janeway, Jr. and Ruslan Medzhitov

KIR: DIVERSE, RAPIDLY EVOLVING RECEPTORS OF INNATE AND ADAPTIVE IMMUNITY, Carlos Vilches and Peter Parham ORIGINS AND FUNCTIONS OF B-1 CELLS WITH NOTES ON THE ROLE OF CD5, Robert Berland and Henry H. Wortis E PROTEIN FUNCTION IN LYMPHOCYTE DEVELOPMENT, Melanie W. Quong, William J. Romanow, and Cornelis Murre

197 217 253 301

LYMPHOCYTE-MEDIATED CYTOTOXICITY, John H. Russell and Timothy J. Ley

SIGNAL TRANSDUCTION MEDIATED BY THE T CELL ANTIGEN x

323

P1: FRK

Annu. Rev. Immunol. 2002.20:621-667. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

February 1, 2002

12:47

Annual Reviews

AR152-FM

CONTENTS

xi

RECEPTOR: THE ROLE OF ADAPTER PROTEINS, Lawrence E. Samelson INTERACTION OF HEAT SHOCK PROTEINS WITH PEPTIDES AND ANTIGEN PRESENTING CELLS: CHAPERONING OF THE INNATE AND ADAPTIVE IMMUNE RESPONSES, Pramod Srivastava CHROMATIN STRUCTURE AND GENE REGULATION IN THE IMMUNE SYSTEM, Stephen T. Smale and Amanda G. Fisher PRODUCING NATURE’S GENE-CHIPS: THE GENERATION OF PEPTIDES FOR DISPLAY BY MHC CLASS I MOLECULES, Nilabh Shastri, Susan

371

Schwab, and Thomas Serwold

395 427

463

THE IMMUNOLOGY OF MUCOSAL MODELS OF INFLAMMATION, Warren Strober, Ivan J. Fuss, and Richard S. Blumberg T CELL MEMORY, Jonathan Sprent and Charles D. Surh

GENETIC DISSECTION OF IMMUNITY TO MYCOBACTERIA: THE HUMAN MODEL, Jean-Laurent Casanova and Laurent Abel ANTIGEN PRESENTATION AND T CELL STIMULATION BY DENDRITIC CELLS, Pierre Guermonprez, Jenny Valladeau, Laurence Zitvogel, Clotilde Th´ery, and Sebastian Amigorena

495 551 581

621

NEGATIVE REGULATION OF IMMUNORECEPTOR SIGNALING, Andr´e Veillette, Sylvain Latour, and Dominique Davidson

669

CPG MOTIFS IN BACTERIAL DNA AND THEIR IMMUNE EFFECTS, Arthur M. Krieg

PROTEIN KINASE Cθ

709 IN

T CELL ACTIVATION, Noah Isakov and Amnon

Altman

RANK-L AND RANK: T CELLS, BONE LOSS, AND MAMMALIAN EVOLUTION, Lars E. Theill, William J. Boyle, and Josef M. Penninger PHAGOCYTOSIS OF MICROBES: COMPLEXITY IN ACTION, David M. Underhill and Adrian Ozinsky

761 795 825

STRUCTURE AND FUNCTION OF NATURAL KILLER CELL RECEPTORS: MULTIPLE MOLECULAR SOLUTIONS TO SELF, NONSELF DISCRIMINATION, Kannan Natarajan, Nazzareno Dimasi, Jian Wang, Roy A. Mariuzza, and David H. Margulies

853

INDEXES Subject Index Cumulative Index of Contributing Authors, Volumes 1–20 Cumulative Index of Chapter Titles, Volumes 1–20

ERRATA An online log of corrections to Annual Review of Immunology chapters (if any, 1997 to the present) may be found at http://immunol.annualreviews.org/

887 915 925

8 Feb 2002

10:7

AR

AR152-21.tex

AR152-21.SGM

LaTeX2e(2001/05/10) P1: GJC 10.1146/annurev.immunol.20.081501.130710

Annu. Rev. Immunol. 2002. 20:669–707 DOI: 10.1146/annurev.immunol.20.081501.130710 c 2002 by Annual Reviews. All rights reserved Copyright °

Annu. Rev. Immunol. 2002.20:669-707. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

NEGATIVE REGULATION OF IMMUNORECEPTOR SIGNALING Andr´e Veillette1–4, Sylvain Latour5 and Dominique Davidson1 1

Laboratory of Molecular Oncology, IRCM, 110 Pine Avenue West, Montr´eal, Qu´ebec, Canada H2W 1R7 2 Department of Medicine, University of Montr´eal, Montr´eal, Qu´ebec, Canada Departments of 3Biochemistry and 4Microbiology and Immunology, McGill University, Montr´eal, Qu´ebec, Canada 5 Unit´e INSERM U429, Hˆopital Necker Enfants-Malades, Paris, France e-mails: [email protected] (A.V.); [email protected] (S.L.); [email protected] (D.D.)

Key Words Csk, phosphatase, adaptor, ITIM, inhibition ■ Abstract Immune cells are activated as a result of productive interactions between ligands and various receptors known as immunoreceptors. These receptors function by recruiting cytoplasmic protein tyrosine kinases, which trigger a unique phosphorylation signal leading to cell activation. In the recent past, there has been increasing interest in elucidating the processes involved in the negative regulation of immunoreceptormediated signal transduction. Evidence is accumulating that immunoreceptor signaling is inhibited by complex and highly regulated mechanisms that involve receptors, protein tyrosine kinases, protein tyrosine phosphatases, lipid phosphatases, ubiquitin ligases, and inhibitory adaptor molecules. Genetic evidence indicates that this inhibitory machinery is crucial for normal immune cell homeostasis.

INTRODUCTION A proper immune response requires the balanced participation of several immune cell types, including T lymphocytes, B lymphocytes, natural killer (NK) cells, dendritic cells, macrophages, mast cells, and other myeloid cells. Over the past fifteen years, significant progress has been made toward understanding the molecular mechanisms that trigger the functional activation of these various cell types and explaining their relative contribution to the immune response. Most notable is our comprehension of the activation of T cells and B cells, which constitute the major components of antigen-specific immunity (1–4). Whereas the mechanisms of antigen recognition by T and B cells are different, the intracellular signaling machinery leading to activation of these two cell types is highly conserved. This machinery is also shared by receptors for the Fc portion 0732-0582/02/0407-0669$14.00

669

4 Feb 2002

11:58

Annu. Rev. Immunol. 2002.20:669-707. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

670

AR

AR152-21.tex

VEILLETTE

¥

AR152-21.SGM

LATOUR

¥

LaTeX2e(2001/05/10)

P1: GJC

DAVIDSON

of immunoglobulins (FcRs), found on a wide range of hemopoietic cells (5, 6). Furthermore, it is used by natural cytotoxicity receptors (NCRs) on NK cells and by activating variants of inhibitory receptors like killer inhibitory receptors (KIRs) and Ly49 (7, 8). On the basis of their conserved mechanism of signal transduction, the TCR, BCR, FcRs, NCRs, and the activating variants of inhibitory receptors are referred to as immunoreceptors (Table 1). As a counterbalance to this activating apparatus, immune cells possess a complex and organized machinery aimed at restricting the duration and/or intensity of cell activation. This machinery involves receptors, kinases, phosphatases, adaptor molecules, and others. It is crucial for normal immune homeostasis, as exemplified by the observation that alterations of its components can result in autoimmune diseases, inflammatory disorders, and lymphoid malignancies.

TABLE 1 Immunoreceptors Receptors

Distribution

a) Classical immunoreceptors TCR T-cells

Ligands

Ligand-binding subunits

ITAM subunits

CD3γ ,δ,ε TCRζ Igα, Igβ

Src kinases

Syk kinases

Lck, FynT

Zap-70, (Syk)

Lyn, Blk, FynT, Lck Lyn, FynT, c-Src, cYes

Syk

BCR

B-cells

Soluble Ag

FcεRI

Mast cells, basophils, monocytes

IgE

αβ γδ mIgM, mIgD, mIgG FcεRIα

Fcγ RI

Macrophages, myeloid cells Monocytes, myeloid cells NK cells, myeloid cells, mast cells Human NK cells

IgG

Fcγ RI

FcεRIγ

Hck, Lyn

Syk

IgG

Fcγ RIIA

Fcγ RIIA

c-Fgr, Hck, Lyn

Syk

IgG

Fcγ RIII

Lck, Lyn

Syk

? (target cells)

NKp44, NKp46, NKp30

FcεRIγ TCRζ FcεRIγ TCRζ DAP-12

Lck, Lyn

Syk, Zap-70

Class I MHC

DAP-12

?

Syk, Zap-70

HLA-E

KIR2DS, KIR3DS NKG2C,E

DAP-12

?

Syk, Zap-70

Class I MHC ?

Ly49D,H ?

DAP-12 FcεRIγ

? ?

Syk ?

?

?

FcεRIγ

?

?

?α vβ 3

?

?FcεRIγ

?

?

? ?CD47

? ?

DAP-12 DAP-12

? ?

? ?

?

?

DAP-12

?

?

Fcγ RIIA Fcγ RIII NCRs

b) Non-classical immunoreceptors KIR2DS, Human NK cells KIR3DS CD94/ NK cells NKG2C,E Ly49D,H Mouse NK cells PIR-A Mouse B-cells, myeloid cells ILT1,7,8, Human NK cells, LIR6a B-cells, myeloid cells gp49A ? mouse mast cells, ?NK cells TREM-1, -2 Myeloid cells SIRPβ Macrophages, dendritic cells MDL-1 Monocytes, macrophages

Antigen-MHC

FcεRIγ FcεRIβ

Syk

8 Feb 2002

10:34

AR

AR152-21.tex

AR152-21.SGM

LaTeX2e(2001/05/10)

P1: GJC

INHIBITORY SIGNALING

671

This review covers the current knowledge regarding the negative regulation of immune cell activation. Prior to entering into this topic, however, we briefly summarize the signaling machinery involved in immunoreceptor-mediated cell activation.

Annu. Rev. Immunol. 2002.20:669-707. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

POSITIVE REGULATION OF IMMUNORECEPTOR SIGNALING Even though immunoreceptors lack intrinsic catalytic activity, their engagement results in a rapid protein tyrosine phosphorylation signal (2, 4, 5, 7, 9, 10). This signal is essential for immunoreceptor-mediated function. It is mediated by a short tyrosine-based cytoplasmic motif located in subunits associated with the immunoreceptors (Table 1). This sequence is termed immunoreceptor tyrosinebased activation motif (ITAM), and it consists of YxxL/I-x(6-8)-YxxL/I (where Y is tyrosine, L is leucine, I is isoleucine, and x is any residue). ITAMs function by coordinating the recruitment of three classes of cytoplasmic protein tyrosine kinases (PTKs), namely the Src family, the Syk family, and the Btk family (2, 4, 5, 7, 9). The Src-related PTKs, which include c-Src, Lck, Fyn, Lyn, c-Fgr, Hck, Blk, c-Yes, and Yrk, are responsible for the initiation of immunoreceptor signaling. In response to immunoreceptor engagement, they become activated and phosphorylate the two tyrosines of the ITAMs (3, 10, 11). This phosphorylation triggers a series of signaling events, which leads to cell activation. Src-related enzymes are regulated by tyrosine phosphorylation (Figure 1). Autophosphorylation of a tyrosine in the activation loop (Y394 in Lck) results in an increase in their catalytic activity. Conversely, phosphorylation of a carboxyl-terminal tyrosine (Y505 in Lck) provokes a dramatic decrease in their enzymatic function. This inhibitory phosphorylation is caused by the cytoplasmic PTK Csk. In the case of Lck and possibly Fyn, it is reversed by CD45, a transmembrane protein tyrosine phosphatase (PTP) (12, 13). The phosphorylation of ITAMs by Src kinases allows recruitment of the tandem Src homology 2 (SH2) domain-containing PTKs Syk and Zap-70 (3, 10, 14). Whereas Zap-70 is restricted to T cells and NK cells, Syk is expressed in B cells, mast cells, myeloid cells, macrophages, and platelets. Syk also accumulates in certain T cell subsets such as immature thymocytes and γ δ T cells. The enzymatic activity of Syk and Zap-70 is also regulated via tyrosine phosphorylation. Phosphorylation of a tyrosine residue in the activation loop of the kinase domain (Y520 in Syk) is key to the enzymatic activation of these two molecules. While phosphorylation of this site occurs through autophosphorylation in Syk, it is mediated by Src kinases in the case of Zap-70. Inhibitory phosphorylation sites have also been identified in Syk and Zap-70. In particular, a phosphorylated tyrosine in the so-called linker domain (Y317 in Syk) inhibits the function of Syk family kinases as a result of its capacity to recruit the negative regulator c-Cbl. When activated, Syk family kinases cause the tyrosine phosphorylation of downstream effectors

4 Feb 2002

11:58

Annu. Rev. Immunol. 2002.20:669-707. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

672

AR

AR152-21.tex

VEILLETTE

¥

AR152-21.SGM

LATOUR

¥

LaTeX2e(2001/05/10)

P1: GJC

DAVIDSON

such as the adaptors LAT, SLP-76, and Blnk. They also cooperate with Src family kinases to phosphorylate phospholipase C (PLC)-γ and Vav. The Btk family of PTKs includes Btk, Itk, Tec, Bmx, and an atypical member termed Rlk/Txk (15, 16). Typical Btk kinases are recruited to the plasma membrane by a mechanism that requires prior activation of phosphatidylinositol (PI) 30 kinase by Src and Syk family kinases. When activated, PI 30 kinase generates membrane-bound PI metabolites phosphorylated at the 30 position of the inositol ring. One of these, PI(3,4,5)P3, can bind with high affinity to the pleckstrin homology (PH) domain of Btk kinases and allows their membrane association. Once at the membrane, Btk kinases are activated by tyrosine phosphorylation. In the case of Btk, this phosphorylation occurs in the activation loop (Y551) and the Src homology 3 (SH3) domain (Y223). It is mediated in part by Src kinases and in part by autophosphorylation. Despite the knowledge that Btk family kinases play an important role in immunoreceptor signaling, information regarding the nature of their substrates is limited. However, it seems that phosphorylation of PLC-γ is particularly dependent on proper Btk family function. Proteins that undergo tyrosine phosphorylation in response to immunoreceptor stimulation can be subdivided into two categories: enzymatic effectors and adaptors. These molecules are briefly summarized in Table 2. For more detailed reviews on these topics, the reader is referred to other publications (4, 5, 7, 9, 10, 17–19).

NEGATIVE REGULATION OF IMMUNORECEPTOR SIGNALING In light of the central role played by protein tyrosine phosphorylation in immunoreceptor signaling, inhibition of this event provides an efficient means of interfering with cell activation. Consequently, the inhibitory PTK Csk as well as several PTPs are potent inhibitors of immune cell activation.

Inhibition of Protein Tyrosine Phosphorylation by Kinases CSK: A KEY NEGATIVE REGULATOR OF SRC FAMILY KINASES

Structure, expression, and function Csk is a 50 kilodalton (kDa)-cytoplasmic PTK expressed in all cell types (10, 11, 20). However, it accumulates in greater amounts in hemopoietic cells. The structure of Csk closely resembles that of the Src family. It includes, from the amino-terminus to the carboxyl-terminus: (a) an SH3 domain, (b) a Src homology 2 (SH2) domain, and (c) a catalytic domain. Csk has the exquisite ability to phosphorylate the inhibitory tyrosine of Src kinases, thereby suppressing their activity (20). In keeping with this notion, ablation of Csk expression in the mouse germ line results in hyperactive Src kinases, severe neural tube defects, and lethality at day 9–10 of embryonic development (21, 22). Moreover, overexpression of Csk has been found to inhibit signal transduction initiated by Src kinases in a variety of cell types.

Activation (Y551 and Y223 in Btk) Binding of PTB domain-containing effectors

PTKs

Coordinate signaling PTKs

PTKs

Src family

ITAM-containing subunits

Syk family

Btk family

Binding of SH2 domain-containing partners Binding of SH2 domain-containing partners

Adaptor Adaptor PTKs Ubiquitin ligases Adaptors Exchange factors for Rho, Rac and cdc42 Adaptors Adaptors Adaptor Transmembrane adaptor Inhibitory receptors

Cas

Fyb/SLAP-130

Fak family

Cbl family

Vav family

SLP-76 family

Dok family

Shc

LAT

ITIM-containing receptors

?CD148 SHP-1

Binds PLC-γ , GADS-SLP-76, Grb2-Sos, PI 30 K Bind SHP-1, SHP-2, SHIP-1 and SHIP-2

PTP-PEST

?

SHP-1 c-Cbl

Cbl-b

?PEP

PTP-PEST

?

Bind SHIP-1, Grb2-Sos

Bind Ras-GAP, SHIP-1, Csk, Nck

?CD148 PTP-PEST

INHIBITORY SIGNALING

Binding of SH2 domain-containing partners

Binding of SH2 domain-containing partners

Bind Vav, Nck, GADS, Grb2, Btk kinases, PLC-γ , PI 30 K, Fyb/SLAP-130, HPK-1

Actin reorganization Jnk activation

Syk family, Src family, SLP-76 family, immunoreceptors, Vav-1, PI 30 K

Focal adhesion-associated proteins

Binds SLP-76 Actin reorganization

Focal adhesion-associated proteins

?

LaTeX2e(2001/05/10)

Binding of SH2 domain-containing partners

Activation

?Activation

Activation Binding of SH2 domain-containing partners

Dephosphorylates PI(3,4,5)P3 and I(1,3,4,5)P4 Hydrolyse PI(4,5)P2 to produce IP3 and DAG

SHIP-1 ?PTEN

SHP-1 c-Cbl ?PEP

AR152-21.SGM

Binding of SH2 domain-containing partners

Binding of SH2 domain-containing partners

Activation

50 inositol phosphatase Enzymes

SHIP

Phosphorylate PLC-γ

Phosphorylate LAT, SLP-76, Blnk, PLC-γ , ?ITAMs (Syk) ?Recruit Vav, PLC-γ

SHP-1 c-Cbl

AR152-21.tex

Phospholipase C-γ

Activation (Y520 in Syk) Inhibition (Y317 and Y519 in Syk) ?Binding of SH2 domain-containing partners

Bind Syk kinases

Csk-PEP SHP-1 CD45 c-Cbl

Inhibitory mechanisms

AR

Binding of SH2 domain-containing partners

Phosphorylate ITAMs, Syk family, Btk family Phosphorylate SHIP-1, PLC-γ , Fyb/SLAP-130, Fak family, c-Cbl, Vav-1, Dok family, Shc Phosphorylate ITIMs

Targets

11:58

Activation (Y394 in Lck) Inhibition (Y505 in Lck)

Functions

Substrates

Impact of tyrosine phosphorylation

TABLE 2 Substrates involved in immunoreceptor signaling

Annu. Rev. Immunol. 2002.20:669-707. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

4 Feb 2002 P1: GJC

673

4 Feb 2002

11:58

Annu. Rev. Immunol. 2002.20:669-707. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

674

AR

AR152-21.tex

VEILLETTE

¥

AR152-21.SGM

LATOUR

¥

LaTeX2e(2001/05/10)

P1: GJC

DAVIDSON

The involvement of Csk in the negative regulation of immunoreceptor signaling was first revealed by studies showing that augmented expression of Csk in an antigen-specific T cell line (BI-141) resulted in a marked inhibition of TCRinduced protein tyrosine phosphorylation and interleukin-2 (IL-2) production (23). This effect was rescued by expression of an Src kinase mutant in which the inhibitory tyrosine was replaced by phenylalanine, implying that the repressive effect was mediated by inactivation of Src-related PTKs. Csk overexpression was also shown to inhibit signaling and phagocytosis triggered by high-affinity Fc receptors for IgG (Fcγ Rs) in a macrophage cell line (RAW 264.7) (24). By opposition, though, Csk was found to have a much weaker effect on FcεRI-induced signaling and degranulation in a basophil leukemia cell line (RBL 2H3) (25). Likewise, it was unable to repress BCR-induced signal transduction in a B cell line (WEHI-231) (L. M. L. Chow, A.Veillette, unpublished results). Moreover, elimination of Csk expression failed to alter significantly BCR signaling in the DT-40 B cell line (26). Thus, Csk is a potent inhibitor of immunoreceptor signaling in T cells and macrophages, but not in B cells and mast cells. While the basis for this divergence is not known, it may reflect differences in the absolute need of Src kinases for the initiation of immunoreceptor signaling in these various cell types. Syk has the capacity to induce ITAM tyrosine phosphorylation and trigger immunoreceptor signaling in the absence of functional Src-related PTKs (10). Hence, in B cells and mast cells, endogenous Syk molecules may rescue immunoreceptor-induced signals in the absence of active Src kinases. Alternatively, it is possible that B cells and mast cells lack molecules that are required for the inhibitory influence of Csk. Csk is also essential for proper T cell development (27, 28). Conditional gene ablation experiments using the Cre-lox approach showed that T cell maturation can take place independently of TCR or MHC expression in thymocytes lacking Csk (normally, T cell differentiation is strictly dependent on MHC-driven signal transduction events). Such an effect is eliminated in Csk-deficient thymocytes lacking Lck and Fyn, the two Src kinases expressed in thymocytes, indicating that it is mediated by hyperactive Src kinases. Presumably, the absence of Csk causes TCR-independent activation of Src kinases, thereby leading to unrestricted T cell maturation. It is not known whether a similar effect exists in Csk-deficient immature B cells. Based on the available data, the absence of Csk may have a less dramatic impact on B cell differentiation. Regulation by protein-protein interactions In view of the importance of Csk for the inhibition of Src kinases, there has been significant interest in elucidating its regulation. Structure-function analyses showed that, in addition to its catalytic activity, the SH3 and SH2 domains of Csk are absolutely critical for its capacity to inhibit TCR-mediated signal transduction (29). Because SH3 and SH2 domains mediate protein-protein interactions, it is therefore likely that the function of Csk requires associations with other cellular proteins. Through a yeast two-hybrid screen, it was revealed that the Csk SH3 region interacts with PEP, a proline-rich PTP expressed only in hemopoietic cells (30).

8 Feb 2002

10:36

AR

AR152-21.tex

AR152-21.SGM

LaTeX2e(2001/05/10)

P1: GJC

Annu. Rev. Immunol. 2002.20:669-707. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

INHIBITORY SIGNALING

675

Approximately 50%–80% of all PEP molecules are constitutively associated with Csk in these cells. By sequence homology, it was also determined that Csk associates by way of its SH3 domain with PTP-PEST, another member of PEP family expressed in a broad range of cell types (31). Therefore, it appears that the Csk SH3 domain is responsible for constitutive associations with proline-enriched protein tyrosine phosphatases (Table 3). The purpose of these interactions will be discussed below. In contrast, the SH2 region of Csk is involved in inducible interactions with tyrosine phosphorylated molecules (Table 3). Several years ago, a provocative report suggested that this domain might allow direct binding of Csk to Src kinases, thereby providing a direct means of recruiting the inhibitory kinase to its targets (32). However, this finding was not confirmed by others (33). More recent studies have indicated that the Csk SH2 domain binds to molecules that are substrates of Src-like PTKs, thus providing an indirect way of recruiting Csk in the vicinity of activated Src kinases. Most notably, two groups have identified a novel transmembrane molecule termed PAG (phosphoprotein associated with GEMs) or Cbp (Csk-binding protein), which appears to be the major Csk SH2 domain-binding protein in immune cells (34, 35). PAG/Cbp (hereafter named PAG) is an ∼80–90-kDa type III transmembrane protein expressed ubiquitously. It predominantly accumulates in lipid rafts, probably due to palmitylation of two cysteines located in the membraneproximal region of its cytoplasmic domain. PAG possesses a short extracellular domain (16–18 amino acids), a single transmembrane region, as well as a long cytoplasmic domain bearing ten potential sites of tyrosine phosphorylation. One of these tyrosines, Y317 in human PAG, was reported to be the binding site for Csk. In addition to allowing recruitment of Csk near lipid rafts, binding of PAG to the Csk SH2 region increases the kinase activity of Csk (36). There is compelling evidence that PAG tyrosine phosphorylation is mediated by Src-related PTKs. In transfected Cos cells, PAG can be phosphorylated by Src kinases, but not by Syk kinases (35). Moreover, PAG tyrosine phosphorylation is absent in Lck-deficient Jurkat T cells. Finally, a small quantity of Fyn can be co-immunoprecipitated with PAG in T cells. Apparently, this association is phosphotyrosine-independent. On this basis, it may involve the Fyn SH3 domain. PAG is constitutively tyrosine phosphorylated and associated with approximately 20% of Csk molecules in resting T cells (35, 37). Once T cells are activated by anti-CD3 antibodies, PAG reportedly undergoes rapid dephosphorylation and dissociates from Csk. This loss of the PAG-Csk interaction is transient, correlating with the induction of overall protein tyrosine phosphorylation. It may be instrumental in allowing the Src kinase activation required for the initiation of TCR signaling. While this notion remains hypothetical for now, it is supported by the observation that overexpression of PAG in Jurkat T cells provoked an inhibition of TCR-mediated signaling events (35). Taking these observations into account, the following model can be proposed (Figure 2). In resting T cells, Src kinases in lipid rafts cause constitutive

Transmembrane Ubiquitous adaptor

Adaptor

Adaptor

PTK

Adaptor

Transmembrane Lymphocytes adaptor ?Receptor

PAG

Dok-1

Dok-3

Fak

Paxillin

SIT

Membrane

SH2

SH2

Focal adhesions SH2

Inhibits BCR signaling

Inhibits BCR signaling

SHP-2, Grb2

Fak, PTP-PEST, others

Dephosphorylates activating tyrosine Src kinases ?Regulates clavage furrows

?Recruits Csk to the membrane

Organizes focal adhesions

Paxillin, Src kinases, Organizes focal Grb2, others adhesions

SHIP-1

Ras-GAP, SHIP-1, Nck

pY354 and pY381 PSTPIP, PSTPIP2 (mouse)

pY168 (human)

Ptyr

Ptyr

Ptyr

Ptyr

Recruits Csk to lipid rafts

Dephosphorylates Shc, Fak, Pyk2 and Cas

LaTeX2e(2001/05/10)

Primitive hemopoietic cells Cytosol

Ubiquitous

SH2

SH2

Focal adhesions SH2

Cytosol ?Membrane

Cytosol ?Membrane

Fyn

Shc, paxillin, Cas, PSTPIP

AR152-21.SGM

Ubiquitous

B-cells, macrophages

High in hemopoietic cells

pY317 (human)

P2

¥

SH2

SH3

Dephosphorylates activating tyrosine Src kinases, as well as Zap-70

LATOUR

Membrane Lipid rafts

Cytosol

?c-Cbl

¥

Ubiquitous High in hemopoietic cells

P1

AR152-21.tex

PTP-HSCF PTP

PTP

PTP-PEST

SH3

Effects

VEILLETTE

Cytosol ?Nucleus

Localization

Other associated proteins

AR

Hemopoietic cells

PTP

PEP

Expression

Functions

Binding domain in Csk Binding site

13:37

Binding proteins

676

TABLE 3 Csk-binding proteins

Annu. Rev. Immunol. 2002.20:669-707. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

14 Feb 2002 P1: GJC

DAVIDSON

8 Feb 2002

10:36

AR

AR152-21.tex

AR152-21.SGM

LaTeX2e(2001/05/10)

P1: GJC

Annu. Rev. Immunol. 2002.20:669-707. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

INHIBITORY SIGNALING

677

tyrosine phosphorylation of PAG and the formation of PAG-Csk complexes. PAGassociated Csk helps maintain Src kinases in a hypoactive state. Once the TCR is engaged, PAG undergoes transient dephosphorylation by an as-yet-undetermined mechanism. This dephosphorylation provokes a loss of lipid raft-associated Csk, which in turn allows an increase in Src kinase activity and permits the initiation of T cell activation. Further insights into this model will undoubtedly be provided by the elucidation of the mechanism of TCR-induced PAG dephosphorylation. It is conceivable that Csk-associated PEP and PTP-PEST are responsible for this phenomenon. Alternatively, other PTPs such as SHP-1 or CD45 may be involved. It is also possible that TCR stimulation triggers PAG dephosphorylation by inhibiting or sequestering the PTKs responsible for phosphorylating PAG. In addition to the Csk-binding site, PAG bears other tyrosines in its cytoplasmic region that are likely to be phosphorylated. This suggests that PAG may recruit additional SH2 domain-containing molecules to lipid rafts. If this is the case, it will be important to determine the identity of these polypeptides and to assess whether they are positive or negative regulators of cell signaling. It is conceivable that, in the absence of Csk binding, PAG mediates a positive signal in lipid rafts. The creation and analysis of PAG-deficient cells should help address this possibility. Other molecules have the ability to bind the Csk SH2 domain. In particular, two members of the Dok family of adaptors, Dok-1 and Dok-3, were reported to associate with Csk by this mechanism (38, 39). These polypeptides are efficient inhibitors of immunoreceptor signaling and recruit not only Csk but also the 50 inositol phosphatase SHIP-1 and Ras-GTPase-activating protein (Ras-GAP). The Dok family is further discussed below. Additionally, Csk can associate with SIT (SHP2-interacting transmembrane adaptor protein), a transmembrane adaptor molecule exclusively expressed in lymphocytes (40, 41). Overexpression of SIT in Jurkat T cells inhibits TCR-induced activation of NFAT (nuclear factor of activated T cells), and this inhibitory impact correlated with the ability of SIT to interact with Csk (40). Csk can also associate through its SH2 domain with paxillin, tensin, and focal adhesion kinase (FAK), three focal adhesion–associated proteins (42). However, because focal adhesions do not exist in immune cells, the physiological relevance of these associations for immune cell regulation is unclear. Lastly, Csk is able to bind in an SH2 domain-dependent manner to PTP-HSCF, a third member of the PEP family selectively expressed in primitive hemopoietic cells (43, 44). This finding is somewhat unexpected, since the other members of the PEP family, PEP and PTP-PEST, bind to Csk via its SH3 domain. Contrary to its two relatives, PTP-HSCF can undergo tyrosine phosphorylation in response to Src kinase activation. Phosphorylation of two tyrosines in its carboxyl-terminal region is responsible for binding to the Csk SH2 region. As is the case for the Csk-PEP association (see below), the interaction of Csk with PTP-HSCF seems to augment the capacity of Csk to inhibit Src kinases. Regulation by phosphorylation By opposition to Src-like PTKs, Csk is not prominently regulated by phosphorylation. Nevertheless, it was reported that Csk can

4 Feb 2002

11:58

678

AR

AR152-21.tex

VEILLETTE

¥

AR152-21.SGM

LATOUR

¥

LaTeX2e(2001/05/10)

P1: GJC

DAVIDSON

Annu. Rev. Immunol. 2002.20:669-707. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

be phosphorylated on serine by cAMP-dependent protein kinase (45). Seemingly, this phosphorylation occurs at serine 364 in the catalytic loop of Csk and results in an increase in Csk activity. It might explain the previously described ability of cAMP treatment to inhibit T cell activation. There is also some indication that Csk can undergo phosphorylation on tyrosine residues (46). However, the biological relevance of this phosphorylation is not clear because its detection requires treatment of cells with potent PTP inhibitors. CHK: A SECOND MEMBER OF THE CSK FAMILY The Csk family of PTKs contains a second member named Chk (also previously identified as Ctk, Lsk, Ntk, Hyl, MATK, and Batk) (11). Chk is expressed in brain and hemopoietic cells. Like Csk, it is capable of phosphorylating the inhibitory tyrosine of Src kinases in vitro. Furthermore, it can substitute for Csk and restore regulation of Src-like PTKs in Csk-deficient fibroblasts (47). In contrast, Chk is inefficient at inhibiting TCR signaling (47). Whereas the basis of this phenomenon is still uncertain, it may relate to the observation that Chk is incapable of binding to PEP and PTP-PEST (30, 31). It may also indicate that Chk is restricted to a cellular compartment distinct from that of TCR-regulated Src kinases.

Protein Tyrosine Phosphatases Involved in the Negative Regulation of Immunoreceptor Signaling Several PTPs have been implicated in the inhibition of immunoreceptor-mediated signal transduction. They include cytoplasmic PTPs such as SHP-1, PEP, and PTP-PEST, as well as receptor-like PTPs such as CD45 and CD148. SH2 DOMAIN-CONTAINING PROTEIN TYROSINE PHOSPHATASES SHP-1 and SHP-2 are two related PTPs containing tandem amino-terminal SH2 domains, a central phosphatase domain, and a carboxyl-terminal extension of undetermined function (Figure 3) (48–51). While SHP-1 is restricted to hemopoietic cells and epithelial cells, SHP-2 is expressed ubiquitously. SHP-1 and SHP-2 intervene in cell signaling as a consequence of binding of their tandem SH2 domains to tyrosine phosphorylated docking sites. Whereas SHP-1 is typically an inhibitor of signal

Figure 3 Primary structure of phosphatases implicated in the negative regulation of immunoreceptor signaling. See text for details.

4 Feb 2002

11:58

AR

AR152-21.tex

AR152-21.SGM

LaTeX2e(2001/05/10)

P1: GJC

Annu. Rev. Immunol. 2002.20:669-707. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

INHIBITORY SIGNALING

679

transduction, SHP-2 is usually a positive regulator of cell signaling. For this reason, we focus our discussion on SHP-1. Interested readers are referred elsewhere for reviews on SHP-2 (51–54). The activity of SHP-1 is primarily regulated by conformational modification (48–51). Currently available data suggest that SHP-1 is maintained inactive by way of an intramolecular interaction between its tandem SH2 domains and its phosphatase domain. This view is supported by the finding that deletion of the aminoterminal SH2 module of SHP-1 (but not the carboxyl-terminal SH2 domain) results in strong enzymatic activation of SHP-1 (55–57). Likewise, binding of tyrosine phosphorylated peptides to the SH2 domains of SHP-1 provokes an increase (approximately 10-fold) in the catalytic activity of SHP-1 (55). Lastly, analysis of the crystal structure of the related enzyme SHP-2 demonstrates that, in the inactive state, the amino-terminal SH2 domain of SHP-2 interacts with the PTP domain, thereby interfering with substrate binding (58). Role of SHP-1 in immunoreceptor signaling A significant clue regarding the role of SHP-1 in immune cell homeostasis was provided by the finding that its gene is mutated in motheaten (me) and viable motheaten (mev) mice (59–61). In both cases, the defect in SHP-1 protein expression is caused by a point mutation of a splice site in the shp-1 gene. me mice die at approximately 3 weeks of age, usually due to severe hemorrhagic pneumonitis. mev mice succumb from a similar disease at a later age (9–12 weeks). The difference in the severity of the disease between these two mouse strains is likely explained by the low residual SHP-1 activity found in mev. me and mev mice exhibit severe abnormalities in multiple hemopoietic lineages. In particular, they demonstrate a marked accumulation of myeloid and monocytic cells in several tissues. While this observation provides a clear indication of the importance of SHP-1 in hemopoietic cells, the complexity of the mouse phenotype has made it arduous to determine precisely which signaling pathways are affected by SHP-1 deficiency. Experiments using other systems have established that SHP1 actually inhibits signaling through cytokine receptors, chemokine receptors, integrins, as well as immunoreceptors. The role of SHP-1 in cytokine, chemokine, and integrin signaling is not reviewed here [see Zhang et al. (50) for a review on the topic]. me (or mev)-derived thymocytes exhibit augmented TCR-induced protein tyrosine phosphorylation and IL-2 production (62, 63). In addition, crossing of mev with various TCR transgenic mice has indicated that these mice have increased positive and negative selection (64–66). These alterations are likely due to an intrinsic T cell defect, as they were also observed in T cells from transgenic mice expressing a dominant-negative SHP-1 (cysteine 453-to-serine 453 mutant) (64, 65). An inhibitory role of SHP-1 in TCR signaling was also revealed by transfections in established T cell lines. Overexpression of dominant-negative SHP-1 in Jurkat T cells and 3L2 T cells increased TCR-induced signals (67). Moreover, expression of a membrane-targeted version of SHP-1 (created through addition

4 Feb 2002

11:58

Annu. Rev. Immunol. 2002.20:669-707. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

680

AR

AR152-21.tex

VEILLETTE

¥

AR152-21.SGM

LATOUR

¥

LaTeX2e(2001/05/10)

P1: GJC

DAVIDSON

of class I MHC extracellular and transmembrane sequences) provoked a decrease in TCR-elicited inositol phosphate production and NFAT activation in Jurkat T cells (68). A similar effect was seen with lipid raft-targeted variants of SHP-1 (69, 70). Similar evidence is available regarding the inhibitory impact of SHP-1 on BCR signaling. B cells from motheaten mice exhibit augmented BCR-induced proliferation, protein tyrosine phosphorylation, and MAPK activation (71). This phenomenon was observed both in ex vivo purified B cells and in an established B cell line obtained from me mice (72). Furthermore, B cells isolated from me mice expressing a hen egg lysosyme (HEL)-specific BCR transgene demonstrate clonal deletion with lower valency forms of antigen (soluble HEL) (73). These forms of antigen are usually unable to cause deletion of normal B cells. It is likely that the altered BCR signaling in me mice is intrinsic to B cells, as similar defects were observed in a B cell line (K46 B) expressing a dominant-negative form of SHP-1 (74). In combination, these results establish that SHP-1 is a critical negative regulator of immunoreceptor signaling in T cells and B cells. Whether a similar role exists in other immune cells awaits clarification. Regulation of SHP-1 during immunoreceptor signaling The most significant advance in understanding the regulation of SHP-1 was provided by the observation that it associates with a specific motif found in the intracytoplasmic region of various receptors with inhibitory activity (Table 4). This motif is termed immunoreceptor tyrosine-based inhibitory motif (ITIM) (75–78). It consists of the sequence (I/V/L/S)xYxx(L/V), where I is isoleucine, V is valine, L is leucine, S is serine, Y is tyrosine, and x is any residue. Typically, immune receptors that associate with SHP-1 possess two or more ITIMs. The prevailing model indicates that ITIM-containing receptors are tyrosine phosphorylated in response to immunoreceptor engagement, as a result of the action of Src kinases (75–78). This phosphorylation creates high-affinity binding sites for the SH2 domains of SHP-1 (and occasionally SHP-2, SHIP-1, or SHIP-2). As a result, SHP-1 is recruited to the plasma membrane and activated through conformational modification. Activated SHP-1 dephosphorylates various immunoreceptor-regulated substrates, thereby inhibiting cell activation. This model is clearly valid for some ITIM-containing receptors for which the ligand provokes juxtaposition to an activating receptor (Table 4). These include KIRs, CD94/NKG2A, Ly49, CD22, CD72, ILTs, SHPS-1/SIRP-α and gp49B1. Because of space limitation, the reader is referred to other reviews for further discussions on these receptors (7, 8, 51, 78–82). However, in the case of ITIMcontaining receptors like PECAM-1, CEACAM-1, SIGLEC5-7, and PIR-B, they may not undergo tyrosine phosphorylation in response to immunoreceptor engagement because it is unclear how their ligand would allow sufficient approximation to an activating receptor to trigger ITIM phosphorylation. Hence, even though these receptors may be capable of inhibiting immunoreceptor signaling when artificially co-aggregated with activating receptors, they may not be physiological

4 Feb 2002

11:58

AR

AR152-21.tex

AR152-21.SGM

LaTeX2e(2001/05/10)

P1: GJC

Annu. Rev. Immunol. 2002.20:669-707. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

INHIBITORY SIGNALING

681

negative regulators of immunoreceptor signaling. Rather, they may participate in the inhibition of other signaling pathways. The ability of SHP-1 deficiency, or expression of dominant-negative SHP-1, to enhance TCR- and BCR-induced responses established that SHP-1 is a negative regulator of immunoreceptor signaling (48–51). However, these observations did not address the mechanism(s) by which SHP-1 intersects with the antigen receptor signaling machinery, in the absence of deliberate co-engagement of an ITIM-containing receptor. In B cells, it seems that SHP-1 can be recruited to CD22 or CD72 even in the absence of their co-aggregation with the BCR. However, it is also conceivable that SHP-1 is enlisted by molecules that lack ITIMs. In keeping with this idea, it was reported that SHP-1 can be found in a complex with the BCR in resting B cells. This association is decreased following BCR stimulation and may involve associations with cytoskeletal constituents (71). Likewise, SHP-1 can be co-immunoprecipitated with the TCR complex both in resting and in activated T cells (63). SHP-1 can also associate with several intracellular polypeptides participating in immunoreceptor signaling such as Syk family kinases, SLP-76-related adaptors, PI 30 kinase, Vav, and Grb2. While the biological relevance of these various interactions remains to be elucidated, these findings suggest that SHP-1 can also be regulated by intracellular molecules. Targets of SHP-1 in immunoreceptor signaling Analyses of protein tyrosine phosphorylation in SHP-1-deficient immune cells, co-immunoprecipitation experiments, as well as in vitro dephosphorylation studies, have indicated that SHP-1 is likely to act at several levels in the immunoreceptor signaling cascade (48–51). For instance, SHP-1 was reported to dephosphorylate Src kinases, ITAMs, Syk kinases, adaptors such as SLP-76, effectors like Vav and PI 30 kinase, as well as receptors such as CD19. Moreover, there is strong indication that SHP-1 can dephosphorylate ITIM-containing receptors, thereby providing a potential mechanism of autoregulation. Unfortunately, currently available technologies do not permit determination of which, if any, of these substrates are the most relevant targets of SHP-1. PEP-RELATED PROTEIN TYROSINE PHOSPHATASES

PEP PEP is the prototype of the PEP family of nonreceptor PTPs, which also contains PTP-PEST and PTP-HSCF (Figure 4) (Table 3) (83, 84). It possesses an amino-terminal phosphatase domain, in addition to a long carboxyl-terminal extension bearing several proline-rich sequences. PEP is expressed exclusively in hemopoietic cells, including T cells, B cells, and macrophages. In T cells, it is found both in thymocytes and in mature T cells (84). Its expression can be further enhanced by TCR stimulation. There is some controversy regarding the intracellular distribution of PEP. It was initially reported that PEP is located in the nucleus (85). However, this finding was not confirmed by others (30). Seemingly, most of PEP is distributed in the cytoplasm. It is possible, nevertheless, that a fraction of PEP translocates to the nucleus under certain conditions.

9p

CD72

H,M

Lectin-like

Ig

B-cells

1

SHP-1

SHP-1

SHP-1, SHP-2

?SHP-1

SHP-1, SHP-2

gp49A

SIRPβ

NKG2C,E

Ly49D,H

LaTeX2e(2001/05/10)

M only

10 (mouse)

?1

1–3

Macrophages, dendritic 3 cells Mast cells, NK cells 2

gp49B1

Ig

H,M

20p13

Hemopoietic cells

SHP-1, SHP-2

NK cells, CD8+ T-cells 2

SHPS-1 (SIRPα)

Lectin-like

SHP-1, SHP-2

NK cells, CD8+ T-cells 1

Scavenger-like T-cells, B1 B-cells

H,M

Lectin-like

H,M

12p13

NKG2A (with CD94)

M only

KIR3DS

SHP-1

NK cells, CD8+ T-cells 2

Ig

6 (mouse)

Ly49A-C,E-G,I

Ig

¥

H only

Ig

LATOUR

H,M

19q13.4

KIR3DL

H only

AR152-21.SGM

SIGLEC family 19q13.1-3 (CD22, CD33, etc.) CD5 11q13

19q13.4

Inhibition of BCR, FcεRI, Fcγ RIIA signaling HLA-C Inhibition of natural killing, ADCC, T-cell-mediated cytotoxicity HLA-A,-B Inhibition of natural killing, ADCC, T-cell-mediated cytotoxicity Class I MHC Inhibition of natural killing, ADCC, T-cell-mediated cytotoxicity HLA-E (H), Inhibition of natural killing, b ADCC, T-cell-mediated Qal (M) cytotoxicity Sialic acid Inhibition of BCR signaling (CD22) ?CD72 Inhibition of TCR and BCR signaling CD47 Inhibition of phagocytosis (Fcγ RI) α vβ 3 integrin Inhibition of FcεRI signaling Inhibition of NK cell-mediated cytotoxicity ?CD5 Inhibition of BCR signaling

IgG

Functions

AR152-21.tex

KIR2DL

SHIP-1, SHIP-2 Fcγ RIII, Fcγ RIIA SHP-1 KIR2DS

Activating homologues Ligands

B-cells, myeloid cells, 1 mast cells NK cells, CD8+ T-cells 2

Associated # of ITIMs phosphatases

¥

Igc

Expression pattern

VEILLETTE

a) Known to inhibit immunoreceptor signaling H,Mb Fcγ RIIB 1q23-24a

Chr. location Species Class

682

AR

Receptors

11:58

TABLE 4 ITIM-containing inhibitory receptors

Annu. Rev. Immunol. 2002.20:669-707. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

4 Feb 2002 P1: GJC

DAVIDSON

H: human; M: mouse.

Ig: immunoglobulin-like molecule.

b

c

1q21

17

12p13

H,?

H,? H H,?

Lectin-like Myeloid cells Ig B-cells Ig Myeloid cells, B-cells, T-cells, NK cells Ig Hemopoietic cells

Ig

Hemopoietic and epithelial cells Hemopoietic cells

Broad

Myeloid cells

2

1 2 3

2

2

2

2

SHP-1

SHP-1 ? ?

SHP-1, SHP-2

SHP-1, SHP-2

SHP-1, SHP-2

SHP-1

CMRF-35A

?PIRLβ

?

? ? ?

?

? ? ?

LaTeX2e(2001/05/10)

Unless specified, chromosomal location corresponds to human gene.

SPAP1

c) Unknown status LLIR BXMAS1 family CMRF-35H

19q13.4

Ig

Ig

Ig

ILT1,7,8 LIR6a SHP-1, SHP-2 PIR-A

SHP-1

AR152-21.SGM

a

H,M

H only

17q23

PECAM-1 (CD31)

H,M

LAIR-1

7q22

PIRLα

M only

H,M

7 (mouse)

PIR-B

H only

Inhibition of FcεRI signaling (when coligated) Class I MHC Inhibition of FcγRIII, TCR, BCR, (ILT2 and 4) FcγRI signaling (when coligated) ? Inhibition of FcεRI and BCR signaling (when coligated) ? Inhibition of FcγRIIA signaling (when coligated) PECAM-1, Inhibition of TCR signaling α vβ 3 integrin (when coligated) CEACAM-1 Inhibition of TCR signaling (when coligated) EpCAM Inhibition of NK cell- and T-cell-mediated cytotoxicity, and BCR signaling (when coligated)

?

AR152-21.tex

CEACAM-1 (CD66) 19q13.2

19q13.4

ILTs/LIRs

SHP-1

AR

Lectin-like Myeloid cells, mast cells, 1 NK cells Myeloid cells, B-cells, 2—4 Igc NK cells, T-cells Ig Myeloid cells, B-cells 3

11:58

b) Possibly inhibit immunoreceptor signaling H,Mb MAFA 12p12-13a

Annu. Rev. Immunol. 2002.20:669-707. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

4 Feb 2002 P1: GJC

INHIBITORY SIGNALING

683

4 Feb 2002

11:58

Annu. Rev. Immunol. 2002.20:669-707. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

684

AR

AR152-21.tex

VEILLETTE

¥

AR152-21.SGM

LATOUR

¥

LaTeX2e(2001/05/10)

P1: GJC

DAVIDSON

Figure 4 Primary structure of the members of the PEP family of protein tyrosine phosphatases. See text for details. The locations of the various motifs involved in protein-protein interactions are indicated.

PEP is constitutively associated with Csk (30). This interaction involves the first proline-rich region in the carboxyl-terminal domain of PEP (P1: PPPLPERTPESFIV; residues essential for the interaction are underlined) (Figure 4) and the SH3 domain of Csk (86). The high affinity and specificity of the Csk-PEP association is dictated by the proline-rich core of P1, as well as by two hydrophobic residues located carboxyl-terminal to this core (isoleucine 625 and valine 626). Most probably, these two residues make additional stabilizing contacts with the Csk SH3 domain. Overexpression studies in a T cell line (BI-141) have suggested that, like Csk, PEP is a potent negative regulator of T cell activation (87). Similar results were obtained when PEP was transiently overexpressed in Jurkat T cells (88). In BI-141 cells, the inhibitory effect of PEP requires not only the phosphatase activity of PEP, but also the sequence (P1) mediating its interaction with Csk. Therefore, PEP seems to cooperate with Csk to inhibit TCR signaling, by a mechanism that requires their physical interaction. Biochemical studies and substrate-trapping experiments have provided evidence that PEP acts proximally in the TCR signaling cascade, by dephosphorylating the positive regulatory site of Src kinases (Y394 in Lck), Zap-70, but not the ITAMs (87). The Csk-PEP complex constitutes a very efficient mechanism to inactivate Src kinase-mediated signaling, as a result of its dual ability to phosphorylate the inhibitory tyrosine of Src kinases (through Csk) and dephosphorylate their positive regulatory site (through PEP). It is also possible that PEP has additional substrates in T cells. In support of this idea, one study reported that PEP is associated constitutively with c-Cbl in human T cells and that overexpression of PEP inhibits c-Cbl tyrosine phosphorylation (84). More work remains to be done to determine the role of PEP in normal T cell physiology. The creation and analysis of PEP-deficient mice, or mice expressing dominant-negative variants of PEP, should help in this matter. The role of PEP in other immune cell types is also nebulous. It was reported that reduction of PEP

8 Feb 2002

10:38

AR

AR152-21.tex

AR152-21.SGM

LaTeX2e(2001/05/10)

P1: GJC

INHIBITORY SIGNALING

685

Annu. Rev. Immunol. 2002.20:669-707. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

expression in a B cell line (WEHI-231) through antisense technology resulted in a decrease in BCR-induced apoptosis (89). Such a finding raised the possibility that PEP may be a positive regulator of some BCR-elicited responses. PTP-PEST Unlike PEP, PTP-PEST is expressed broadly in nonhemopoietic and hemopoietic cells (90, 91). Higher levels are found in immune cells, including thymocytes, splenic T cells and B cells, NK cells, and mast cells (31, 92). PTPPEST physically associates with several signal transduction molecules, including the adaptor molecule Shc, the focal adhesion proteins paxillin, Hic-5, Cas, CasL (or HEF-1), and Sin, the cleavage furrow-associated protein PSTPIP, and Csk (31, 92– 98). These various interactions occur via sequences positioned in the noncatalytic region of PTP-PEST (Figure 4). The association between PTP-PEST and Shc is further enhanced by BCR engagement or phorbol esters (92). Studies performed in fibroblasts have indicated that PTP-PEST is implicated in the negative regulation of integrin-induced cell migration and spreading, possibly as a result of its capacity to dephosphorylate focal adhesion proteins such as Cas and FAK (99, 100). More recently, the possible involvement of PTP-PEST in the regulation of immunoreceptor signaling was examined (92). Through overexpression studies and antisense experiments in a B cell line (A20), evidence was provided that PTP-PEST is an efficient negative regulator of antigen receptorinduced cytokine production. Similar results were obtained in a T cell line (Jurkat). Biochemical studies and structure-function analyses indicated that the inhibitory impact of PTP-PEST correlates with its ability to dephosphorylate a selective set of tyrosine phosphorylation substrates, including Shc, Cas, Pyk2, and FAK, and to prevent activation of the Ras-MAPK signaling cascade. Surprisingly, binding of Csk is not required for the inhibitory effect of PTP-PEST on BCR signaling. Taken together, these data provide a strong indication that PTP-PEST is a bona fide negative regulator of immunoreceptor signaling in T cells and B cells. Unlike SHP-1 and PEP, it influences a selective subset of substrates in the immunoreceptor signaling cascade (Figure 5). Given this property, it is attractive to speculate that PTP-PEST could modify, in a qualitative rather than quantitative manner, the outcome of antigen receptor stimulation. It may prevent some biological outcomes of antigen receptor stimulation such as cytokine secretion, proliferation, or differentiation, while favoring others like anergy or apoptosis. Future studies, including the characterization of PTP-PEST-deficient lymphocytes, should help address this possibility. CD45 is a transmembrane PTP expressed on all nucleated hemopoietic cells (13, 101, 102). It is abundant and constitutes up to 10% of all membrane proteins in T cells and B cells. CD45 possesses an extracellular domain with variable composition and structure, due to alternative splicing of several exons (4, 5, 6, and 7) and differential glycosylation. There is no known ligand for the extracellular region of CD45, with the possible exceptions of CD22 and galectin that recognize sugar determinants in CD45. In its intracytoplasmic region, CD45

CD45

4 Feb 2002

11:58

Annu. Rev. Immunol. 2002.20:669-707. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

686

AR

AR152-21.tex

VEILLETTE

¥

AR152-21.SGM

LATOUR

¥

LaTeX2e(2001/05/10)

P1: GJC

DAVIDSON

bears two PTP domains, D1 and D2. Whereas D1 has classical catalytic features, D2 may be enzymatically inert. The activity of CD45 is regulated by dimerization of its extracellular domain (103, 104). By creating EGF receptor-CD45 chimeras, it was found that dimerization of CD45 results in a strong inhibition of its PTP activity. Based on the crystal structure of another receptor-like PTP (PTPα), this dimerization is presumed to trigger contacts between a wedge sequence located in the membrane-proximal region of the cytoplasmic domain of one CD45 molecule and the catalytic domain (D1) of another CD45 polypeptide. Even though there is no evidence that CD45 dimerization is induced by physiological ligands, this dimerization might occur spontaneously, with an efficiency that is inversely proportional to the size of the CD45 extracellular domain. CD45 can regulate multiple pathways in immune cells, including signal transduction through immunoreceptors, integrins, and cytokine receptors (12, 101, 102). The function of CD45 in immunoreceptor signaling is mostly stimulatory, due to its ability to dephosphorylate the inhibitory tyrosine of Src kinases. However, CD45 can also have an inhibitory effect on immunoreceptor-elicited signal transduction (12, 13). For instance, TCR-mediated signal transduction can be inhibited by antibody-mediated co-aggregation of the TCR complex with CD45. Moreover, some T cell and B cell lines lacking CD45 exhibit elevated baseline levels of phosphotyrosine and hyperactive Src kinases. Given the aptitude of CD45 to dephosphorylate both the inhibitory and the activating tyrosines of Src kinases, these two functions need to be differentially regulated in cells. One possibility is that the effect of CD45 is dictated by the state of activation of Src kinases at the moment they encounter CD45. If Src kinases are inactive due to phosphorylation of their inhibitory tyrosine, a productive interaction with CD45 would favor activation. If, however, they are phosphorylated at their activating tyrosine, CD45 would have an inhibitory impact. It is conceivable that temporally regulated sequestration of CD45 from Src kinases is needed to allow phosphorylation of the activating tyrosine, once the inhibitory tyrosine has been dephosphorylated. In agreement with this scenario, CD45 seems to be rapidly excluded from the immune synapse in antigen receptor-stimulated T cells and B cells (102). Although much of the work to date has focused on Src-like enzymes, there may be additional substrates for CD45 in immune cells. These include the TCRassociated ζ chain, CD22, Ras-GAP, Vav, as well as Jak kinases. However, the biological relevance of these observations is still uncertain. OTHER PROTEIN TYROSINE PHOSPHATASES

CD148 CD148 (also named HPTP-η or DEP-1) is a receptor-like PTP expressed in a variety of cells, including hemopoietic cells (105). Within immune cells, it accumulates most abundantly in myeloid cells and macrophages. Expression is low in resting T cells but is strongly induced upon antigen receptor–induced cell activation. To date, no ligand has been identified for the extracellular domain of

4 Feb 2002

11:58

AR

AR152-21.tex

AR152-21.SGM

LaTeX2e(2001/05/10)

P1: GJC

INHIBITORY SIGNALING

687

Annu. Rev. Immunol. 2002.20:669-707. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

CD148. Overexpression studies in Jurkat T cells have shown that CD148 is able to inhibit TCR-induced tyrosine phosphorylation of LAT and PLC-γ 1, as well as induction of intracellular calcium fluxes and MAPK activation (106, 107). It has little effect on tyrosine phosphorylation of Zap-70, SLP-76, c-Cbl, and Pyk-2. Given that LAT is located in lipid rafts, these observations raise the possibility that CD148 may be involved in terminating signal transduction in rafts. Unfortunately, though, investigators were unable to show that CD148 accumulates in lipid rafts (106). Characterization of the physiological role of CD148 should be helped by the generation of CD148-deficient mice. Other phosphatases Other nonreceptor phosphatases can inhibit immunoreceptor signaling. They include PTP-H1, HePTP, and dual-specificity phosphatases (DSPs) like PAC-1 and VHR. In most cases, the inhibitory effect of these phosphatases was documented in transient transfection assays (108–111). For HePTP and DSPs, it is proposed that they inhibit T cell activation by selectively inactivating MAPKs. Whether these various phosphatases play a physiological role in immune cell inhibition needs elucidation.

Lipid Phosphatases Involved in the Inhibition of Immunoreceptor Signaling Phosphorylation of membrane-associated lipids plays an important role in immunoreceptor signaling. First, the hydrolysis of PI(4, 5)P2 by activated PLC-γ leads to production of inositol trisphosphate (IP3) and diacylglycerol (DAG), which stimulate the calcium, protein kinase C (PKC), and Ras-MAPK pathways. Second, activation of PI 30 kinase permits phosphorylation of PI at the 30 position, thereby generating PI(3,4,5)P3, PI(3,4)P2, and I(1,3,4,5)P4. These lipids are implicated in membrane recruitment and activation of PH domain-containing effectors such as Btk kinases and the serine/threonine-specific protein kinase Akt. As a corollary, evidence is mounting that dephosphorylation of lipids provides an efficient means of inhibiting immunoreceptor signaling. In particular, the SH2 domain-containing 50 inositol phosphatase SHIP-1 is now recognized as a potent negative regulator of immunoreceptor-mediated signal transduction in B cells, mast cells, and macrophages. Moreover, PTEN, a 30 inositol phosphatase, has been implicated in dampening immune functions in T cells and B cells. SHIP-1, A 50 INOSITOL PHOSPHATASE EXPRESSED IN HEMOPOIETIC CELLS

SHIP-1 is a 145-kDa polypeptide expressed in most, if not all, hemopoietic cells (112–115). It contains an amino-terminal SH2 domain, a central lipid phosphatase domain, and a long carboxyl-terminal region, bearing at least two tyrosine phosphorylation sites (NPxY motifs; where N is asparagine, P is proline, x is any residue, and Y is tyrosine) and several proline-rich sequences (Figure 3). Different versions of SHIP-1 can be produced as a result of alternative initiation of transcription or translation, differential splicing, or proteolysis. SHIP-1 specifically dephosphorylates

4 Feb 2002

11:58

Annu. Rev. Immunol. 2002.20:669-707. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

688

AR

AR152-21.tex

VEILLETTE

¥

AR152-21.SGM

LATOUR

¥

LaTeX2e(2001/05/10)

P1: GJC

DAVIDSON

lipids at the 50 position of the inositol ring. Interestingly, lipids also need to be phosphorylated at the 30 position (by PI 30 kinase) before they can be targeted by SHIP-1. Consequently, SHIP-1 can act only on PI(3,4,5)P3 and I(1,3,4,5)P4. Characterization of SHIP-1-deficient mice has provided firm evidence that SHIP-1 is a crucial negative regulator of several processes in hemopoietic cells (116). Animals lacking SHIP-1 die early in life, due to myeloid cell infiltration of several organs, especially lungs. B cell maturation is also defective (117–119). These phenotypic anomalies are caused by enhanced signaling through a variety of receptors (116–123). These include receptors for cytokines, growth factors, and chemokines, as well as immunoreceptors. The increased signaling is seemingly due to augmented levels of PI(3,4,5)P3 and I(1,3,4,5)P4 (118). Importance of SHIP-1 in the negative regulation of immunoreceptor signaling SHIP-1 plays a crucial role in the inhibition of BCR, FcεRI and Fcγ R signaling. B cells from mice lacking SHIP-1 have been reported to develop abnormally. In the spleen, there is a reduced number of immature transitional B cells (IgMhiIgDlo and IgMhiIgDhi) and an increased quantity of mature B cells (IgMloIgDhi) (117–119). Moreover, reconstitution of the B cell compartment in autotransplanted irradiated mice is accelerated (118). Studies of isolated B cells from SHIP-1-deficient mice have also shown that BCR-induced proliferation is increased (119). This was attributed at least in part to a decrease in BCR-triggered cell death (118). There are also enhanced antigen receptor–elicited calcium fluxes, activation of Erk and Akt, as well as induction of B cell activation markers such as CD86 and CD69. Lastly, BCR-induced elevation of PI(3,4,5)P3 production is augmented. Thus, in combination, these observations indicate that B cell maturation and B cell activation are increased in the absence of SHIP-1. One caveat to these studies is that multiple immune cell lineages are affected in SHIP-1-deficient animals. Thus, it is possible that part of the alterations of BCR functions is indirect, due to defects in other cell types. However, this possibility is rendered less likely by the characterization of a SHIP-1-deficient variant of the DT-40 B cell line. These cells also demonstrate an augmentation of BCR-induced calcium mobilization (124), caused by increased calcium release from intracellular stores (125). In addition, membrane association of the PH domain-containing PTK Btk and the activity of the PH domain-containing protein kinase Akt are increased (121, 126). SHIP-1 also inhibits signaling through the high-affinity receptor for IgE (FcεRI). Mast cells derived from mice deficient in SHIP-1 exhibit more pronounced FcεRI-induced calcium mobilization, Erk activation, and degranulation (123). Interestingly, whereas granule release triggered by FcεRI requires stimulation by multimerized IgE in normal mast cells, it can be induced by monomeric IgE in SHIP-1-deficient cells. Given the demonstrated involvement of Btk in FcεRIinduced calcium fluxes and degranulation, it is probable that the manifestations of SHIP-1 deficiency in mast cells are caused in part by sustained Btk activation. In macrophages, SHIP-1 participates in the negative regulation of phagocytosis mediated by Fc receptors for IgG (Fcγ Rs). In support of this idea, SHIP-1-deficient

8 Feb 2002

10:40

AR

AR152-21.tex

AR152-21.SGM

LaTeX2e(2001/05/10)

P1: GJC

Annu. Rev. Immunol. 2002.20:669-707. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

INHIBITORY SIGNALING

689

macrophages were found to demonstrate elevated Fcγ R-triggered phagocytosis (127). Furthermore, overexpression of wild-type SHIP-1 in a macrophage cell line (RAW LR5) inhibited phagocytosis mediated by Fcγ Rs, whereas expression of a catalytically inactive version of SHIP-1 augmented this response. It is worthy of note that the inhibitory effect of SHIP-1 on phagocytosis was blocked by an inactivating mutation in the SHIP SH2 domain, implying that it involved the binding and recruitment of SHIP-1 to a tyrosine phosphorylated docking molecule. It is surprising that T cell development and activation are not detectably affected by the lack of SHIP-1. Perhaps, SHIP-1 is not absolutely necessary for the regulation of TCR signaling due to functional redundancy with other lipid phosphatases such as SHIP-2 or PTEN. Alternatively, PI(3,4,5)P3-activated pathways may have a less evident role in the immunoreceptor-mediated development and activation of T cells. Along these lines, recent evidence has suggested that SHIP-1 may be involved in modulating the types of cytokines produced during T cell activation, by interacting with immune cell receptors such as SLAM (128, 129). Regulation of SHIP-1 during immunoreceptor signaling The functional activity of SHIP-1 in immune cells is principally regulated by changes in its intracellular localization. It can be recruited to the plasma membrane by binding of its SH2 domain to tyrosine phosphorylated ITIM-containing receptors. The best known case of ITIM-mediated recruitment of SHIP-1 is that of Fcγ RIIB, a low-affinity receptor for IgG expressed on B cells, mast cells, and macrophages (Table 4). Engagement of Fcγ RIIB by the Fc portion of IgG triggers tyrosine phosphorylation of its single ITIM by Src kinases and recruitment of SHIP-1. SHIP-1 provides a strong inhibitory signal for cell activation, in large part by inhibiting Btk family kinases. For further details on the mechanism of Fcγ RIIB inhibition, the reader is referred to a recent review on this topic (78). Other ITIM-containing receptors such as gp49B1 and PECAM-1 also interact with SHIP-1 in vitro. However, in vivo, these receptors have a preferential affinity for SHP-1. Lastly, it was recently reported that SLAM, a receptor modulating interferon-γ production by activated T cells, physically associates with SHIP-1 via an ITIM-like sequence (128, 129). The increase in BCR, FcεRI, or Fcγ R signaling observed in SHIP-1-deficient cells implies that SHIP-1 can inhibit immunoreceptor signaling in the absence of deliberate co-aggregation of the immunoreceptor with an ITIM-bearing receptor. Intracellular adaptors such as Dok-related polypeptides and Shc are likely involved in this process. The Dok family of adaptors comprises five known members named Dok-1, Dok-2 (also termed Dok-R and FRIP), Dok-3 (also named Dok-L), Dok-4, and Dok-5 (10, 130). These molecules exhibit an amino-terminal PH domain, a central phosphotyrosine-binding (PTB) domain, and a carboxyl-terminal region bearing several sites of tyrosine phosphorylation (Figure 6). Dok-1, Dok-2, and Dok-3 are particularly abundant in hemopoietic cells (39). Whereas Dok-1 accumulates in most hemopoietic cells, Dok-2 is present in T cells, mast cells, and macrophages, but not in B cells. By opposition, Dok-3 abounds in B cells,

8 Feb 2002

10:41

Annu. Rev. Immunol. 2002.20:669-707. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

690

AR

AR152-21.tex

VEILLETTE

¥

AR152-21.SGM

LATOUR

¥

LaTeX2e(2001/05/10)

P1: GJC

DAVIDSON

mast cells, and macrophages, but not in T cells. Dok-4 and Dok-5 are found mainly in nonhemopoietic cells (130). In immune cells, Dok-related polypeptides undergo tyrosine phosphorylation in response to a variety of stimuli, such as immunoreceptor ligation, growth factors, and cytokines (39, 131–136). This phosphorylation triggers SH2 domain-mediated interactions with inhibitory effectors including not only SHIP-1 (Dok-2 and Dok-3; less with Dok-1), but also Csk (Dok-1 and Dok-3) and Ras-GAP (Dok-1 and Dok-2) (Figure 6). Several experiments have established that Dok-related polypeptides are inhibitory molecules, presumably due to their capacity to coordinate the recruitment of these three inhibitory effectors. They inhibit signaling through immunoreceptors, receptor PTKs, and cytokine receptors (39, 131–138). Moreover, Dok-1 and Dok-3 can inhibit transformation by Bcr-Abl and v-Abl, respectively (134, 139). The regulatory role of Dok family adaptors in immunoreceptor signaling has been most firmly documented in B cells (39, 132, 137). Both Dok-1 and Dok-3 undergo prompt tyrosine phosphorylation in response to BCR engagement. Dok-1 overexpression in Ramos B cells also inhibits BCR-induced activation of the c-fos promoter (140). Moreover, by analyzing Dok-1-deficient B cells, it was established that Dok-1 is crucial for the inhibitory effect of Fcγ RIIB on BCR-elicited Erk activation and cell proliferation (137). In the case of Dok-3, enforcement of its expression in A20 B cells was reported to result in diminished BCR-induced IL-2 production, while introduction of a Dok-3 variant that cannot be tyrosine phosphorylated caused an enhancement of antigen receptor–stimulated production of IL-2 (39). Thus, Dok-1 and Dok-3 are implicated in the negative regulation of BCR signaling, presumably through their ability to recruit SHIP-1, Csk, and/or Ras-GAP (Figure 6). The role of Dok-related molecules in the regulation of immunoreceptor signaling in other immune cell types is largely undetermined. Nevertheless, Dok-3 was shown to undergo tyrosine phosphorylation and to associate with SHIP-1 in response to engagement of Fcγ RI on the macrophage cell line J774A (39). Therefore, it is likely that Dok-related polypeptides inhibit immunoreceptor signaling in other cell types as well. Shc is an intracellular adaptor molecule expressed ubiquitously (141). It possesses an amino-terminal PTB domain, a central region with sites of tyrosine phosphorylation, and a carboxyl-terminal SH2 domain. Immunoreceptor engagement triggers tyrosine phosphorylation of Shc, as well as its association with SHIP-1. The binding of SHIP-1 to Shc occurs through a bidentate mechanism, involving the Shc PTB domain and the NPxY motifs of SHIP-1 on the one hand, and the sites of tyrosine phosphorylation of Shc and the SHIP-1 SH2 region on the other hand (112, 113). In B cells, the Shc-SHIP-1 and Fcγ RIIB-SHIP-1 interactions are mutually exclusive, presumably because both complexes solicit the SH2 domain of SHIP-1. Whereas the exact purpose of the Shc-SHIP-1 interaction is not known, one scenario is that Shc may be involved in intracellular targeting of SHIP-1. However, it is also possible that SHIP-1 is implicated in recruiting Shc to certain cellular

4 Feb 2002

11:58

AR

AR152-21.tex

AR152-21.SGM

LaTeX2e(2001/05/10)

P1: GJC

INHIBITORY SIGNALING

691

Annu. Rev. Immunol. 2002.20:669-707. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

locales. In accordance with this last possibility, BCR-triggered Shc tyrosine phosphorylation was reported to be reduced in SHIP-1-deficient DT-40 B cells (142). Obviously, additional studies are required to determine whether Shc has a role in the intracellular targeting of SHIP-1. This is especially important, as Shc is commonly viewed as a positive regulator, rather than an inhibitory molecule. Nonetheless, it is noteworthy that overexpression of Shc in an NK cell line (NKL) was recently reported to inhibit Fcγ RIII-mediated calcium fluxes and cytotoxicity (143). Whether the inhibitory effect of Shc in these cells is mediated by SHIP-1 or other Shc-associated molecules such as PTP-PEST was not clarified. SHIP-2 is the other known SH2 domain-containing 50 inositol phosphatase (144). It has the same structural organization as SHIP-1. Contrary to SHIP-1, SHIP-2 accumulates not only in immune cells, but also in nonhemopoietic cell types. The role of SHIP-2 in immunoreceptor signaling is poorly understood. Like SHIP-1, it can bind Fcγ RIIB and undergo tyrosine phosphorylation in response to ligation of Fcγ RIIB to BCR (145). Interestingly, while expression of SHIP-2 is low or absent in resting B cells, it is strongly augmented in lipopolysaccharide (LPS)stimulated B cells. This induction may explain the residual inhibitory activity of Fcγ RIIB in LPS-stimulated B cells lacking SHIP-1. Through the creation of SHIP-2-deficient mice, it was established that SHIP-2 is not absolutely required for immune cell development (146). Of course, this could be due to functional compensation by SHIP-1. These mice, however, demonstrate increased insulin sensitivity, supporting the notion that SHIP-2 is a negative regulator of cell signaling. SHIP-2

PTEN, A 30 INOSITOL PHOSPHATASE WITH TUMOR SUPPRESSOR ACTIVITY PTEN is a cytoplasmic phosphatase expressed in a variety of cell types including immune cells (112, 147–149). Whereas earlier reports suggested that PTEN is a dual-specificity protein phosphatase, subsequent analyses indicated that it is a lipid phosphatase specific for the 30 position of the inositol ring. It catalyzes the conversion of PI(3,4,5)P3 and PI(3,4)P2 into PI(4,5)P2 and PI(4)P, respectively. Consequently, PTEN antagonizes the effects of PI 30 kinase on cell signaling. Its phosphatase domain is similar in structure to that of protein phosphatases, although it has an enlarged active site to accommodate PI substrates. PTEN also contains a C2 domain that can bind membrane phospholipids in vitro and may be involved in positioning PTEN at the membrane. pten is a classical tumor suppressor gene. Heterozygous mutations of pten have been documented in spontaneous and hereditary human malignancies, including glioblastomas, prostate cancer, breast cancer, as well as T and B cell lymphomas. The tumor suppressor activity of PTEN was further confirmed by the creation of PTEN-deficient mice (150). While a homozygous null PTEN mutation results in early embryonic lethality, a high frequency of tumors is observed in heterozygous mice. The inhibitory effect of PTEN on tumor growth has been ascribed to its capacity to reduce PI 30 kinase metabolites such as PI(3,4,5)P3 and possibly PI(3,4)P2. As a result, membrane recruitment and activation of the anti-apoptotic

4 Feb 2002

11:58

Annu. Rev. Immunol. 2002.20:669-707. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

692

AR

AR152-21.tex

VEILLETTE

¥

AR152-21.SGM

LATOUR

¥

LaTeX2e(2001/05/10)

P1: GJC

DAVIDSON

kinase Akt are diminished. In the absence of PTEN, Akt becomes hyperactivated and presumably protects cells from physiological apoptotic stimuli. This likely explains the enhanced susceptibility of PTEN-deficient cells to malignant transformation. There is indication that PTEN plays a significant role in immune cell regulation. Humans or mice deficient in PTEN have an increased susceptibility to develop lymphoid malignancies (150). Furthermore, heterozygous PTEN-deficient mice acquire a severe autoimmune disorder involving T cells and B cells, apparently caused by defective antigen receptor–induced Fas-dependent apoptosis. In a related way, mice having T cell–specific loss of PTEN (through Cre-lox-regulated recombination) have impaired central and peripheral tolerance, with defects in TCR-induced cell death (151). These alterations are accompanied by an increase in TCR-triggered activation of Akt and Erk. The inhibitory impact of PTEN on TCR-induced activation of Akt and Erk is also supported by transfection experiments in Jurkat T cells (152). Intriguingly, however, an analysis of PTEN expression in parental Jurkat has revealed that the pten gene is mutated and PTEN protein expression is absent in this cell line (153). A similar situation may apply to other lymphoid cell lines. This finding is not surprising, because Jurkat (and other related cell lines) was derived from a malignant human tumor. Consequent to this lack of PTEN expression, the activity of Akt is constitutively augmented in Jurkat cells. Moreover, the extent of membrane association and activity of Itk, a Btk family PTK, are elevated. In addition to suggesting that PTEN might also regulate Btk-related kinases, these findings raise issues regarding the appropriateness of cell lines such as Jurkat to study PI 30 kinase-mediated signal transduction. Since both PTEN and SHIP-1 can reduce the accumulation of PI(3,4,5)P3 in immune cells and prevent activation of Akt and Btk kinases, it is striking that the effects of PTEN and SHIP-1 deficiency in mice are quite distinct. One or more of the following possibilities could explain this difference. First, intracellular effectors other than Akt and Btk-related PTKs may be inactivated selectively by either PTEN or SHIP-1. This may reflect the added ability of PTEN to dephosphorylate PI(3,4)P2, or of SHIP-1 to dephosphorylate I(1,3,4,5)P4. Second, because of differences in cellular localization, PTEN and SHIP-1 may act on different pools of Akt and Btk kinases. Third, PTEN and SHIP-1 may act differentially on Akt and Btk-like PTKs. Based on the available evidence, it would appear that PTEN influences Akt to a greater extent than Btk kinases, while the opposite may be true for SHIP-1. Thus, PTEN would primarily inhibit cell survival, while SHIP-1 would mostly inhibit cell activation and proliferation. Fourth, it is conceivable that the functional differences between the two phosphatases reflect activities that are independent of their lipid phosphatase function. Indeed, PTEN has been reported to have protein phosphatase activity, while SHIP has an adaptor-like function allowing recruitment of Shc, Dok-related molecules, and Grb2. And fifth, the impact of SHIP-1 deficiency in immune cells may be less striking as a consequence of partial compensation by SHIP-2.

4 Feb 2002

11:58

AR

AR152-21.tex

AR152-21.SGM

LaTeX2e(2001/05/10)

P1: GJC

INHIBITORY SIGNALING

693

Annu. Rev. Immunol. 2002.20:669-707. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

Inhibitory Signaling Mediated by Src-Related Protein Tyrosine Kinases Src kinases are crucial positive regulators of immunoreceptor signaling, consequent to their capacity to phosphorylate the ITAMs. However, these PTKs can also have an inhibitory impact on this response. This notion was first exemplified by analyses of mice lacking Lyn, an Src-related PTK involved in BCR signaling (154–156). These animals were found to develop an autoimmune disease, as a consequence of exaggerated BCR-induced responses. Additional studies demonstrated that Lyn is necessary for tyrosine phosphorylation of Fcγ RIIB and CD22, two ITIM-bearing receptors that downregulate BCR signaling by recruiting SHIP-1 and SHP-1, respectively (157–159). In a similar way, more recent data showed that the Src-related kinase c-Fgr is involved in the negative regulation of Fcγ R-mediated phagocytosis in macrophages, by phosphorylating the ITIM-containing receptor SHPS-1 (160). When tyrosine phosphorylated, SHPS-1 recruits SHP-1, a negative regulator of Fcγ R-triggered signal transduction. Thus, depending on the substrates that they are phosphorylating, Src PTKs can either activate or inhibit immunoreceptor-mediated signal transduction. Obviously, these two functions need to be differentially regulated in order to achieve productive cell activation. The exact mechanism involved in this regulation has not yet been elucidated.

Negative Regulation of Immunoreceptor Signaling by the Cbl Family of Ubiquitin Ligases Immunoreceptor signaling can also be inhibited by interfering with the function of critical mediators through selective protein degradation or steric hindrance. Cbl-related polypeptides carry out important inhibitory functions through such mechanisms (161–163). The Cbl family comprises three members named c-Cbl, Cbl-b, and Cbl-3. Their structure is highly conserved, including from the aminoto the carboxyl-terminus: (a) a phosphotyrosine-binding region encompassing a four-helix bundle, an EF hand and a variant SH2 domain, (b) a RING finger with E3 ubiquitin ligase activity, (c) multiple proline-rich domains and sites of tyrosine phosphorylation, and (d) a leucine zipper-like sequence. c-Cbl and Cbl-b, but not Cbl-3, are found in immune cells. The biological importance of the Cbl family was first appreciated by the observation that a modified version of c-Cbl was the transforming element of a retrovirus (Cas NS-1) capable of causing pro-B, pre-B, and myeloid tumors in mice (161, 164, 165). Other oncogenic alleles of c-cbl were subsequently identified in transformed lymphoid cell lines such as 70Z/3, as well as in malignant human lymphoid tumors (161, 166, 167). In all cases studied, the oncogenic c-Cbl lacks the RING finger domain and thus is devoid of ubiquitin ligase activity. These mutants are thought to represent dominant-negative forms of c-Cbl. Several lines of evidence have suggested that c-Cbl and Cbl-b play crucial roles in immunoreceptor signaling (161–163). Both undergo tyrosine phosphorylation

4 Feb 2002

11:58

Annu. Rev. Immunol. 2002.20:669-707. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

694

AR

AR152-21.tex

VEILLETTE

¥

AR152-21.SGM

LATOUR

¥

LaTeX2e(2001/05/10)

P1: GJC

DAVIDSON

in response to stimulation of immunoreceptors, including TCR, BCR, Fcγ Rs, and FcεRI. This phosphorylation is seemingly mediated by the combined actions of Src and Syk family kinases. In addition, c-Cbl has the capacity to associate with molecules involved in immunoreceptor-mediated signal transduction such as Src kinases, Syk kinases, Vav, PI 30 kinase, Crk, and Grb2. Likewise, Cbl-b can interact with Src kinases, Syk kinases, SLP-76, Vav, PLC-γ 1, and PI 30 kinase. The interaction between Cbl-related molecules and Syk/Zap-70 involves the SH2 domain of c-Cbl or Cbl-b, and a conserved tyrosine in the linker domain of Syk (Y317) or Zap-70 (Y292). Most of the other associations are mediated by prolinerich sequences or tyrosine phosphorylation sites in c-Cbl or Cbl-b. C-CBL, A NEGATIVE REGULATOR OF SYK FAMILY KINASES By transfecting wildtype or dominant-negative c-Cbl in various cell lines, including T cells, B cells, macrophages, and mast cells, sound indication has been obtained that c-Cbl is an inhibitor of immunoreceptor signaling (161–163). This inhibitory influence relates at least in part to the capacity of c-Cbl to bind and inactivate Syk and Zap-70. In support of this idea, expression of Syk or Zap-70 mutants that are unable to interact with c-Cbl renders cells refractory to the suppressive effect of c-Cbl (168–170). Moreover, the inhibitory impact of c-Cbl can be eliminated by mutation of its SH2 domain, which mediates the association with Syk and Zap-70. Importantly, it is also abolished by mutation of the RING finger, which confers E3 ubiquitin ligase activity (171). These findings, coupled with the observation that c-Cbl can trigger ubiquitination of Syk and Src kinases (168, 172), have led to the proposition that c-Cbl inhibits immunoreceptor signaling by inducing ubiquitinmediated degradation of Syk and, possibly, Src kinases. However, analyses of c-Cbl-deficient mice have suggested that this may not be the case (172–175). In keeping with the inhibitory role of c-Cbl, these animals exhibit increased cell numbers in thymus, spleen, and lymph nodes. TCRinduced proliferation of thymocytes, as well as positive selection of a class II MHC-restricted transgenic TCR, are also augmented. Lastly, TCR-mediated tyrosine phosphorylation of Zap-70, SLP-76, and LAT is enhanced. Surprisingly, though, these effects are not caused by an increase of Zap-70 amounts. Rather, they are paralleled by a marked elevation of TCR levels on c-Cbl-deficient thymocytes. Taking into consideration the recent report that Zap-70-associated c-Cbl is also able to induce ubiquitination of the TCR complex (176), these results suggest that the increased TCR signaling observed in c-Cbl-deficient mice is due to lack of ubiquitination and degradation of the TCR complex, rather than of Zap-70. Additional mechanisms could be implicated in the inhibitory activity of c-Cbl in immunoreceptor signaling. In particular, c-Cbl could inhibit protein function by steric hindrance. Experiments performed with c-Cbl-deficient DT-40 B cells have nicely supported this idea (177). These analyses indicated that the ability of c-Cbl to inhibit BCR signaling correlates with its capacity to block the BlnkPLC-γ association, thereby preventing adequate PLC-γ tyrosine phosphorylation.

4 Feb 2002

11:58

AR

AR152-21.tex

AR152-21.SGM

LaTeX2e(2001/05/10)

P1: GJC

INHIBITORY SIGNALING

695

Annu. Rev. Immunol. 2002.20:669-707. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

No effect on Syk levels, Syk kinase activity, or BCR levels was reported in this system. Thus, by binding crucial effectors, c-Cbl may functionally sequester these molecules. The capacity of c-Cbl to act at multiple steps and by several mechanisms may help secure its inhibitory impact on immunoreceptor signaling. CBL-B, A NEGATIVE REGULATOR OF VAV-1 SIGNALING The data indicating that Cbl-b is a negative regulator of immunoreceptor signaling are mostly derived from studies of Cbl-b-deficient mice (178, 179). These animals develop a striking phenotype characterized by spontaneous autoimmunity. While no alterations in thymocyte maturation and function are noted, the activation of mature T cells in Cbl-bdeficient mice is severely abnormal. Most notably, TCR-induced proliferation and IL-2 production can occur in the absence of CD28 co-stimulation (normally, these responses require co-engagement of the TCR with CD28). This phenotype is not accompanied by increased levels of TCR expression or Zap-70 hyperphosphorylation, as has been documented for mice lacking the related molecule c-Cbl. Instead, it correlates with a selective increase in the tyrosine phosphorylation and enzymatic activity of the guanine nucleotide exchange factor Vav-1. Hence, lack of Cbl-b may enhance TCR responsiveness by augmenting the activity of Vav-1. In keeping with this suggestion, T cells from Cbl-b-deficient animals have increased TCR-induced cytoskeletal reorganization, a phenomenon mediated by Vav-1 (180). Furthermore, ablation of Cbl-b expression can partially correct the functional defects observed in Vav-1-deficient mice. The exact means by which Cbl-b regulates Vav-1 is not known. Even though Cbl-b can bind Vav-1, it is unlikely that it suppresses Vav-1 by inducing its degradation, since the levels of Vav-1 are not increased in cells lacking Cbl-b. One possibility is that Cbl-b inhibits an upstream activator of Vav-1. Along these lines, it has been shown that Cbl-b can bind and repress the activity of PI 30 kinase, a known activator of Vav-1 (181). Whether this is due to ubiquitin-mediated degradation, steric hindrance, or other effects was not determined. It will be important to determine the basis for the differential specificity of c-Cbl and Cbl-b in inhibitory signaling. While the two molecules have a similar structure and can bind similar sets of proteins, they clearly regulate different components of the immunoreceptor signaling cascade. Furthermore, whereas lack of c-Cbl has a moderate effect on immune cell homeostasis with no apparent clinical manifestations, Cbl-b deficiency leads to severe autoimmunity.

CONCLUSIONS In this review, we have described several mechanisms participating in the negative regulation of immunoreceptor signaling. These inhibitory mechanisms involve receptors, protein tyrosine kinases, protein tyrosine phosphatases, lipid phosphatases, adaptors, and ubiquitin ligases. Additional mechanisms will undoubtedly be uncovered in the future. Thus, the complex and highly regulated apparatus

4 Feb 2002

11:58

Annu. Rev. Immunol. 2002.20:669-707. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

696

AR

AR152-21.tex

VEILLETTE

¥

AR152-21.SGM

LATOUR

¥

LaTeX2e(2001/05/10)

P1: GJC

DAVIDSON

leading to immune cell activation is counterbalanced by an equally sophisticated series of inhibitory mechanisms. A proper balance between these two types of influences is necessary to avoid autoimmunity (due to excess positive signaling or insufficient negative signaling) or immunodeficiency (due to excess negative signaling or insufficient positive signaling). It is interesting that different types of immune cells seem to rely on distinct subsets of inhibitory mechanisms to regulate cell functions. In particular, T cells are mostly reliant on intracellular mechanisms such as protein tyrosine kinases (Csk), protein tyrosine phosphatases (PEP), and ubiquitin ligases. With the exception of CTLA-4, the use of inhibitory receptors is less prevalent. By opposition, other immune cell lineages, including B cells, macrophages, and mast cells, primarily utilize ITIM-containing receptors and their associated phosphatases SHP-1 and SHIP-1. While the basis for this difference is unknown, it may reflect the less stringent requirements needed for activation of cells other than T cells. Indeed, these cell types respond promptly and fully to simple forms of ligands. To prevent undue cell activation, they may require a more elaborate and regulated inhibitory machinery. In contrast, T cells are activated only when they are triggered by proper antigens presented by proper MHC molecules, in the setting of sufficient engagement of co-receptors and co-stimulatory molecules. Since the ability to activate T cells can be easily compromised by altering the function of one of the stimulating components, a more primitive inhibitory machinery may be a sufficient complement to ward off inappropriate cell activation. Studies of the negative regulation of immunoreceptor signaling have also revealed that certain mechanisms inhibit a subset, instead of all, of the immunoreceptor-triggered signal transduction events. For example, PTP-PEST inhibits BCR-induced tyrosine phosphorylation of Shc, FAK, Pyk2, and Cas. Other substrates are not affected. Such a finding suggests that some regulatory mechanisms may be able to prevent a subset of the biological consequences of cell activation. The net impact of their regulation could therefore be modulatory, rather than purely inhibitory. While the existence of this type of mechanism remains to be more firmly established, it is a logical corollary to molecules such as CD28 and ICOS, which modulate positively the profile of cytokine production in activated T cells. Future work should be aimed at addressing this possibility.

ACKNOWLEDGMENTS We thank Mrs. Jeannine Amyot for help with the manuscript. Work in the authors’ laboratories is supported by grants from the National Cancer Institute of Canada, the Canadian Institutes of Health Research and the CANVAC National Centre of Excellence (to A. V.); and the Institut National de la Sant´e et de la Recherche M´edicale and the Association pour la Recherche sur le Cancer (France) (to S. L.). A.V. is a Senior Investigator of the Canadian Institutes of Health Research, while S. L. is a Scientist from the Centre National de la Recherche Scientifique (France).

4 Feb 2002

11:58

AR

AR152-21.tex

AR152-21.SGM

LaTeX2e(2001/05/10)

P1: GJC

INHIBITORY SIGNALING

697

Visit the Annual Reviews home page at www.annualreviews.org

Annu. Rev. Immunol. 2002.20:669-707. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

LITERATURE CITED 1. Weiss A, Littman DR. 1994. Signal transduction by lymphocyte antigen receptors. Cell 76:263–74 2. Tamir I, Cambier JC. 1998. Antigen receptor signaling: integration of protein tyrosine kinase functions. Oncogene 17:1353–64 3. Chan AC, Desai DM, Weiss A. 1994. The role of protein tyrosine kinases and protein tyrosine phosphatases in T cell antigen receptor signal transduction. Annu. Rev. Immunol. 12:555–92 4. Kane LP, Lin J, Weiss A. 2000. Signal transduction by the TCR for antigen. Curr. Opin. Immunol. 12:242–49 5. Turner H, Kinet JP. 1999. Signalling through the high-affinity IgE receptor Fc epsilonRI. Nature 402:B24–30 6. Daeron M. 1997. Fc receptor biology. Annu. Rev. Immunol. 15:203–34 7. Moretta A, Bottino C, Vitale M, Pende D, Cantoni C, Mingari MC, Biassoni R, Moretta L. 2001. Activating receptors and coreceptors involved in human natural killer cell-mediated cytolysis. Annu. Rev. Immunol. 19:197–223 8. Colonna M, Nakajima H, Cella M. 2000. A family of inhibitory and activating Iglike receptors that modulate function of lymphoid and myeloid cells. Semin. Immunol. 12:121–27 9. Kurosaki T. 2000. Functional dissection of BCR signaling pathways. Curr. Opin. Immunol. 12:276–81 10. Latour S, Veillette A. 2001. Proximal protein tyrosine kinases in immunoreceptor signaling. Curr. Opin. Immunol. 13:299– 306 11. Chow LM, Veillette A. 1995. The Src and Csk families of tyrosine protein kinases in hemopoietic cells. Semin. Immunol. 7:207–26 12. Thomas ML, Brown EJ. 1999. Positive and negative regulation of Src-family

13.

14.

15.

16.

17.

18.

19.

20.

21.

22.

membrane kinases by CD45. Immunol. Today 20:406–11 Ashwell JD, D’Oro U. 1999. CD45 and Src-family kinases: and now for something completely different. Immunol. Today 20:412–16 van Oers NS, Weiss A. 1995. The Syk/ZAP-70 protein tyrosine kinase connection to antigen receptor signalling processes. Semin. Immunol. 7:227–36 Yang WC, Collette Y, Nunes JA, Olive D. 2000. Tec kinases: a family with multiple roles in immunity. Immunity 12:373–82 Lewis CM, Broussard C, Czar MJ, Schwartzberg PL. 2001. Tec kinases: modulators of lymphocyte signaling and development. Curr. Opin. Immunol. 13: 317–25 Rudd CE. 1999. Adaptors and molecular scaffolds in immune cell signaling. Cell 96:5–8 Myung PS, Boerthe NJ, Koretzky GA. 2000. Adapter proteins in lymphocyte antigen-receptor signaling. Curr. Opin. Immunol. 12:256–66 Leo A, Schraven B. 2001. Adapters in lymphocyte signalling. Curr. Opin. Immunol. 13:307–16 Nada S, Okada M, MacAuley A, Cooper JA, Nakagawa H. 1991. Cloning of a complementary DNA for a protein-tyrosine kinase that specifically phosphorylates a negative regulatory site of p60c-src. Nature 351:69–72 Nada S, Yagi T, Takeda H, Tokunaga T, Nakagawa H, Ikawa Y, Okada M, Aizawa S. 1993. Constitutive activation of Src family kinases in mouse embryos that lack Csk. Cell 73:1125–35 Imamoto A, Soriano P. 1993. Disruption of the csk gene, encoding a negative regulator of Src family tyrosine kinases, leads to neural tube defects and embryonic lethality in mice. Cell 73:1117–24

4 Feb 2002

11:58

Annu. Rev. Immunol. 2002.20:669-707. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

698

AR

AR152-21.tex

VEILLETTE

¥

AR152-21.SGM

LATOUR

¥

LaTeX2e(2001/05/10)

P1: GJC

DAVIDSON

23. Chow LM, Fournel M, Davidson D, Veillette A. 1993. Negative regulation of T-cell receptor signalling by tyrosine protein kinase p50csk. Nature 365:156–60 24. Suzuki T, Kono H, Hirose N, Okada M, Yamamoto T, Yamamoto K, Honda Z. 2000. Differential involvement of Src family kinases in Fc gamma receptormediated phagocytosis. J. Immunol. 165: 473–82 25. Honda Z, Suzuki T, Hirose N, Aihara M, Shimizu T, Nada S, Okada M, Ra C, Morita Y, Ito K. 1997. Roles of C-terminal Src kinase in the initiation and the termination of the high affinity IgE receptor-mediated signaling. J. Biol. Chem. 272:25753–60 26. Hata A, Sabe H, Kurosaki T, Takata M, Hanafusa H. 1994. Functional analysis of Csk in signal transduction through the B-cell antigen receptor. Mol. Cell. Biol. 14:7306–13 27. Schmedt C, Saijo K, Niidome T, Kuhn R, Aizawa S, Tarakhovsky A. 1998. Csk controls antigen receptor-mediated development and selection of T-lineage cells. Nature 394:901–4 28. Schmedt C, Tarakhovsky A. 2001. Autonomous maturation of alpha/beta T lineage cells in the absence of COOHterminal Src kinase (Csk). J. Exp. Med. 193:815–26 29. Cloutier JF, Chow LM, Veillette A. 1995. Requirement of the SH3 and SH2 domains for the inhibitory function of tyrosine protein kinase p50csk in T lymphocytes. Mol. Cell. Biol. 15:5937–44 30. Cloutier JF, Veillette A. 1996. Association of inhibitory tyrosine protein kinase p50csk with protein tyrosine phosphatase PEP in T cells and other hemopoietic cells. EMBO J. 15:4909–18 31. Davidson D, Cloutier JF, Gregorieff A, Veillette A. 1997. Inhibitory tyrosine protein kinase p50csk is associated with protein-tyrosine phosphatase PTP-PEST in hemopoietic and non-hemopoietic cells. J. Biol. Chem. 272:23,455–62

32. Bougeret C, Delaunay T, Romero F, Jullien P, Sabe H, Hanafusa H, Benarous R, Fischer S. 1996. Detection of a physical and functional interaction between Csk and Lck which involves the SH2 domain of Csk and is mediated by autophosphorylation of Lck on tyrosine 394. J. Biol. Chem. 271:7465–72 33. Gervais FG, Chow LM, Lee JM, Branton PE, Veillette A. 1993. The SH2 domain is required for stable phosphorylation of p56lck at tyrosine 505, the negative regulatory site. Mol. Cell. Biol. 13:7112–21 34. Kawabuchi M, Satomi Y, Takao T, Shimonishi Y, Nada S, Nagai K, Tarakhovsky A, Okada M. 2000. Transmembrane phosphoprotein Cbp regulates the activities of Src-family tyrosine kinases. Nature 404:999–1003 35. Brdicka T, Pavlistova D, Leo A, Bruyns E, Korinek V, Angelisova P, Scherer J, Shevchenko A, Hilgert I, Cerny J, Drbal K, Kuramitsu Y, Kornacker B, Horejsi V, Schraven B. 2000. Phosphoprotein associated with glycosphingolipid-enriched microdomains (PAG), a novel ubiquitously expressed transmembrane adaptor protein, binds the protein tyrosine kinase csk and is involved in regulation of T cell activation. J. Exp. Med. 191:1591– 604 36. Takeuchi S, Takayama Y, Ogawa A, Tamura K, Okada M. 2000. Transmembrane phosphoprotein Cbp positively regulates the activity of the carboxyl-terminal Src kinase, Csk. J. Biol. Chem. 275: 29183–86 37. Torgersen KM, Vang T, Abrahamsen H, Yaqub S, Horejsi V, Schraven B, Rolstad B, Mustelin T, Tasken K. 2001. Release from tonic inhibition of T cell activation through transient displacement of C-terminal Src kinase (Csk) from lipid rafts. J. Biol. Chem. 276:29,313–18 38. Neet K, Hunter T. 1995. The nonreceptor protein-tyrosine kinase CSK complexes directly with the GTPase-activating protein-associated p62 protein in

4 Feb 2002

11:58

AR

AR152-21.tex

AR152-21.SGM

LaTeX2e(2001/05/10)

P1: GJC

INHIBITORY SIGNALING

39.

Annu. Rev. Immunol. 2002.20:669-707. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

40.

41.

42.

43.

44.

45.

46.

cells expressing v-Src or activated c-Src. Mol. Cell. Biol. 15:4908–20 Lemay S, Davidson D, Latour S, Veillette A. 2000. Dok-3, a novel adapter molecule involved in the negative regulation of immunoreceptor signaling. Mol. Cell. Biol. 20:2743–54 Pfrepper KI, Marie-Cardine A, Simeoni L, Kuramitsu Y, Leo A, Spicka J, Hilgert I, Scherer J, Schraven B. 2001. Structural and functional dissection of the cytoplasmic domain of the transmembrane adaptor protein SIT (SHP2-interacting transmembrane adaptor protein). Eur. J. Immunol. 31:1825–36 Marie-Cardine A, Kirchgessner H, Bruyns E, Shevchenko A, Mann M, Autschbach F, Ratnofsky S, Meuer S, Schraven B. 1999. SHP2-interacting transmembrane adaptor protein (SIT), a novel disulfide-linked dimer regulating human T cell activation. J. Exp. Med. 189:1181–94 Sabe H, Hata A, Okada M, Nakagawa H, Hanafusa H. 1994. Analysis of the binding of the Src homology 2 domain of Csk to tyrosine-phosphorylated proteins in the suppression and mitotic activation of cSrc. Proc. Natl. Acad. Sci. USA 91:3984– 88 Cheng J, Daimaru L, Fennie C, Lasky LA. 1996. A novel protein tyrosine phosphatase expressed in lin(lo)CD34(hi)Sca(hi) hematopoietic progenitor cells. Blood 88: 1156–67 Wang B, Lemay S, Tsai S, Veillette A. 2001. SH2 domain-mediated interaction of inhibitory protein tyrosine kinase Csk with protein tyrosine phosphatase-HSCF. Mol. Cell. Biol. 21:1077–88 Vang T, Torgersen KM, Sundvold V, Saxena M, Levy FO, Skalhegg BS, Hansson V, Mustelin T, Tasken K. 2001. Activation of the COOH-terminal Src kinase (Csk) by cAMP-dependent protein kinase inhibits signaling through the T cell receptor. J. Exp. Med. 193:497–507 Joukov V, Vihinen M, Vainikka S,

47.

48.

49.

50.

51.

52.

53.

54.

55.

56.

57.

699

Sowadski JM, Alitalo K, Bergman M. 1997. Identification of csk tyrosine phosphorylation sites and a tyrosine residue important for kinase domain structure. Biochem. J. 322:927–35 Davidson D, Chow LM, Veillette A. 1997. Chk, a Csk family tyrosine protein kinase, exhibits Csk-like activity in fibroblasts, but not in an antigen-specific T-cell line. J. Biol. Chem. 272:1355–62 Van Vactor D, O’Reilly AM, Neel BG. 1998. Genetic analysis of protein tyrosine phosphatases. Curr. Opin. Genet. Dev. 8:112–26 Tonks NK, Neel BG. 1996. From form to function: signaling by protein tyrosine phosphatases. Cell 87:365–68 Zhang J, Somani AK, Siminovitch KA. 2000. Roles of the SHP-1 tyrosine phosphatase in the negative regulation of cell signalling. Semin. Immunol. 12:361–78 Tamir I, Dal Porto JM, Cambier JC. 2000. Cytoplasmic protein tyrosine phosphatases SHP-1 and SHP-2: regulators of B cell signal transduction. Curr. Opin. Immunol. 12:307–15 Feng GS. 1999. Shp-2 tyrosine phosphatase: signaling one cell or many. Exp. Cell Res. 253:47–54 Huyer G, Alexander DR. 1999. Immune signalling: SHP-2 docks at multiple ports. Curr. Biol. 9:R129–32 Stein-Gerlach M, Wallasch C, Ullrich A. 1998. SHP-2, SH2-containing protein tyrosine phosphatase-2. Int. J. Biochem. Cell. Biol. 30:559–66 Pei D, Wang J, Walsh CT. 1996. Differential functions of the two Src homology 2 domains in protein tyrosine phosphatase SH-PTP1. Proc. Natl. Acad. Sci. USA 93:1141–45 Pei D, Lorenz U, Klingmuller U, Neel BG, Walsh CT. 1994. Intramolecular regulation of protein tyrosine phosphatase SH-PTP1: a new function for Src homology 2 domains. Biochemistry 33:15483– 93 Townley R, Shen SH, Banville D,

4 Feb 2002

11:58

700

58.

Annu. Rev. Immunol. 2002.20:669-707. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

59.

60.

61.

62.

63.

64.

65.

66.

AR

AR152-21.tex

VEILLETTE

¥

AR152-21.SGM

LATOUR

¥

LaTeX2e(2001/05/10)

P1: GJC

DAVIDSON

Ramachandran C. 1993. Inhibition of the activity of protein tyrosine phosphate 1C by its SH2 domains. Biochemistry 32:13414–18 Hof P, Pluskey S, Dhe-Paganon S, Eck MJ, Shoelson SE. 1998. Crystal structure of the tyrosine phosphatase SHP-2. Cell 92:441–50 Tsui HW, Siminovitch KA, de Souza L, Tsui FW. 1993. Motheaten and viable motheaten mice have mutations in the haematopoietic cell phosphatase gene. Nat. Genet. 4:124–29 Shultz LD, Schweitzer PA, Rajan TV, Yi T, Ihle JN, Matthews RJ, Thomas ML, Beier DR. 1993. Mutations at the murine motheaten locus are within the hematopoietic cell protein-tyrosine phosphatase (Hcph) gene. Cell 73:1445–54 Kozlowski M, Mlinaric-Rascan I, Feng GS, Shen R, Pawson T, Siminovitch KA. 1993. Expression and catalytic activity of the tyrosine phosphatase PTP1C is severely impaired in motheaten and viable motheaten mice. J. Exp. Med. 178:2157– 63 Lorenz U, Ravichandran KS, Burakoff SJ, Neel BG. 1996. Lack of SHPTP1 results in src-family kinase hyperactivation and thymocyte hyperresponsiveness. Proc. Natl. Acad. Sci. USA 93:9624–29 Pani G, Fischer KD, Mlinaric-Rascan I, Siminovitch KA. 1996. Signaling capacity of the T cell antigen receptor is negatively regulated by the PTP1C tyrosine phosphatase. J. Exp. Med. 184:839–52 Zhang J, Somani AK, Yuen D, Yang Y, Love PE, Siminovitch KA. 1999. Involvement of the SHP-1 tyrosine phosphatase in regulation of T cell selection. J. Immunol. 163:3012–21 Plas DR, Williams CB, Kersh GJ, White LS, White JM, Paust S, Ulyanova T, Allen PM, Thomas ML. 1999. Cutting edge: the tyrosine phosphatase SHP-1 regulates thymocyte positive selection. J. Immunol. 162:5680–84 Johnson KG, LeRoy FG, Borysiewicz

67.

68.

69.

70.

71.

72.

73.

74.

LK, Matthews RJ. 1999. TCR signaling thresholds regulating T cell development and activation are dependent upon SHP-1. J. Immunol. 162:3802–13 Plas DR, Johnson R, Pingel JT, Matthews RJ, Dalton M, Roy G, Chan AC, Thomas ML. 1996. Direct regulation of ZAP-70 by SHP-1 in T cell antigen receptor signaling. Science 272:1173–76 Musci MA, Beaves SL, Ross SE, Yi T, Koretzky GA. 1997. Surface expression of hemopoietic cell phosphatase fails to complement CD45 deficiency and inhibits TCR-mediated signal transduction in a Jurkat T cell clone. J. Immunol. 158: 1565–71 Kosugi A, Sakakura J, Yasuda K, Ogata M, Hamaoka T. 2001. Involvement of SHP-1 tyrosine phosphatase in TCRmediated signaling pathways in lipid rafts. Immunity 14:669–80 Su MW, Yu CL, Burakoff SJ, Jin YJ. 2001. Targeting Src homology 2 domaincontaining tyrosine phosphatase (SHP-1) into lipid rafts inhibits CD3-induced T cell activation. J. Immunol. 166:3975–82 Pani G, Kozlowski M, Cambier JC, Mills GB, Siminovitch KA. 1995. Identification of the tyrosine phosphatase PTP1C as a B cell antigen receptor-associated protein involved in the regulation of B cell signaling. J. Exp. Med. 181:2077–84 Nadler MJS, Chen B, Anderson JS, Wortis HH, Neel BG. 1997. Proteintyrosine phosphatase SHP-1 is dispensable for FcgammaRIIB-mediated inhibition of B cell antigen receptor activation. J. Biol. Chem. 272:20038–43 Cyster JG, Goodnow CC. 1995. Protein tyrosine phosphatase 1C negatively regulates antigen receptor signaling in B lymphocytes and determines thresholds for negative selection. Immunity 2:13–24 Dustin LB, Plas DR, Wong J, Hu YT, Soto C, Chan AC, Thomas ML. 1999. Expression of dominant-negative src-homology domain 2-containing protein tyrosine phosphatase-1 results in increased Syk

4 Feb 2002

11:58

AR

AR152-21.tex

AR152-21.SGM

LaTeX2e(2001/05/10)

P1: GJC

INHIBITORY SIGNALING

75.

76.

Annu. Rev. Immunol. 2002.20:669-707. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

77.

78. 79. 80.

81.

82.

83.

84.

85.

tyrosine kinase activity and B cell activation. J. Immunol. 162:2717–24 Vivier E, Daeron M. 1997. Immunoreceptor tyrosine-based inhibition motifs. Immunol. Today 18:286–91 Bolland S, Ravetch JV. 1999. Inhibitory pathways triggered by ITIM-containing receptors. Adv. Immunol. 72:149–77 Cambier JC. 1997. Inhibitory receptors abound? Proc. Natl. Acad. Sci. USA 94:5993–95 Ravetch JV, Lanier LL. 2000. Immune inhibitory receptors. Science 290:84–89 Lanier LL. 1998. NK cell receptors. Annu. Rev. Immunol. 16:359–93 Tomasello E, Blery M, Vely E, Vivier E. 2000. Signaling pathways engaged by NK cell receptors: double concerto for activating receptors, inhibitory receptors and NK cells. Semin. Immunol. 12:139– 47 Castells MC, Klickstein LB, Hassani K, Cumplido JA, Lacouture ME, Austen KF, Katz HR. 2001. gp49B1-alpha(v)beta3 interaction inhibits antigen-induced mast cell activation. Nat. Immunol. 2:436–42 Daheshia M, Friend DS, Grusby MJ, Austen KF, Katz HR. 2001. Increased severity of local and systemic anaphylactic reactions in gp49B1-deficient mice. J. Exp. Med. 194:227–34 Matthews RJ, Bowne DB, Flores E, Thomas ML. 1992. Characterization of hematopoietic intracellular protein tyrosine phosphatases: description of a phosphatase containing an SH2 domain and another enriched in proline-, glutamic acid-, serine-, and threonine-rich sequences. Mol. Cell. Biol. 12:2396–405 Cohen S, Dadi H, Shaoul E, Sharfe N, Roifman CM. 1999. Cloning and characterization of a lymphoid-specific, inducible human protein tyrosine phosphatase, Lyp. Blood 93:2013–24 Flores E, Roy G, Patel D, Shaw A, Thomas ML. 1994. Nuclear localization of the PEP protein tyrosine phosphatase. Mol. Cell. Biol. 14:4938–46

701

86. Gregorieff A, Cloutier JF, Veillette A. 1998. Sequence requirements for association of protein-tyrosine phosphatase PEP with the Src homology 3 domain of inhibitory tyrosine protein kinase p50(csk). J. Biol. Chem. 273:13,217–22 87. Cloutier JF, Veillette A. 1999. Cooperative inhibition of T-cell antigen receptor signaling by a complex between a kinase and a phosphatase. J. Exp. Med. 189:111– 21 88. Gjorloff-Wingren A, Saxena M, Williams S, Hammi D, Mustelin T. 1999. Characterization of TCR-induced receptor-proximal signaling events negatively regulated by the protein tyrosine phosphatase PEP. Eur. J. Immunol. 29:3845– 54 89. Hasegawa K, Yajima H, Katagiri T, Ogimoto M, Arimura Y, Mitomo K, Mashima K, Mizuno K, Yakura H. 1999. Requirement of PEST domain tyrosine phosphatase PEP in B cell antigen receptorinduced growth arrest and apoptosis. Eur. J. Immunol. 29:887–96 90. Yang Q, Co D, Sommercorn J, Tonks NK. 1993. Cloning and expression of PTP-PEST. A novel, human, nontransmembrane protein tyrosine phosphatase. J. Biol. Chem. 268:17,650 91. Charest A, Wagner J, Shen SH, Tremblay ML. 1995. Murine protein tyrosine phosphatase-PEST, a stable cytosolic protein tyrosine phosphatase. Biochem. J. 308 (Pt 2):425–32 92. Davidson D, Veillette A. 2001. PTPPEST, a scaffold protein tyrosine phosphatase, negatively regulates lymphocyte activation by targeting a unique set of substrates. EMBO J. 20:3414–26 93. Nishiya N, Iwabuchi Y, Shibanuma M, Cote JF, Tremblay ML, Nose K. 1999. Hic-5, a paxillin homologue, binds to the protein-tyrosine phosphatase PEST (PTPPEST) through its LIM 3 domain. J. Biol. Chem. 274:9847–53 94. Cot´e JF, Charest A, Wagner J, Tremblay ML. 1998. Combination of gene targeting

4 Feb 2002

11:58

702

95.

Annu. Rev. Immunol. 2002.20:669-707. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

96.

97.

98.

99.

100.

101.

102.

AR

AR152-21.tex

VEILLETTE

¥

AR152-21.SGM

LATOUR

¥

LaTeX2e(2001/05/10)

P1: GJC

DAVIDSON

and substrate trapping to identify substrates of protein tyrosine phosphatases using PTP-PEST as a model. Biochemistry 37:13128–37 Shen Y, Schneider G, Cloutier JF, Veillette A, Schaller MD. 1998. Direct association of protein-tyrosine phosphatase PTP-PEST with paxillin. J. Biol. Chem. 273:6474–81 Garton AJ, Burnham MR, Bouton AH, Tonks NK. 1997. Association of PTPPEST with the SH3 domain of p130cas; a novel mechanism of protein tyrosine phosphatase substrate recognition. Oncogene 15:877–85 Garton AJ, Flint AJ, Tonks NK. 1996. Identification of p130(cas) as a substrate for the cytosolic protein tyrosine phosphatase PTP-PEST. Mol. Cell. Biol. 16:6408–18 Charest A, Wagner J, Jacob S, McGlade CJ, Tremblay ML. 1996. Phosphotyrosine-independent binding of SHC to the NPLH sequence of murine protein-tyrosine phosphatase-PEST. Evidence for extended phosphotyrosine binding/phosphotyrosine interaction domain recognition specificity. J. Biol. Chem. 271:8424– 29 Angers-Loustau A, Cot´e JF, Charest A, Dowbenko D, Spencer S, Lasky LA, Tremblay ML. 1999. Protein tyrosine phosphatase-PEST regulates focal adhesion disassembly, migration, and cytokinesis in fibroblasts. J. Cell. Biol. 144: 1019–31 Garton AJ, Tonks NK. 1999. Regulation of fibroblast motility by the protein tyrosine phosphatase PTP-PEST. J. Biol. Chem. 274:3811–18 Alexander DR. 2000. The CD45 tyrosine phosphatase: a positive and negative regulator of immune cell function. Semin. Immunol. 12:349–59 Penninger JM, Irie-Sasaki J, Sasaki T, Oliveira-dos-Santos AJ. 2001. CD45: new jobs for an old acquaintance. Nat. Immunol. 2:389–96

103. Majeti R, Bilwes AM, Noel JP, Hunter T, Weiss A. 1998. Dimerization-induced inhibition of receptor protein tyrosine phosphatase function through an inhibitory wedge. Science 279:88–91 104. Desai DM, Sap J, Schlessinger J, Weiss A. 1993. Ligand-mediated negative regulation of a chimeric transmembrane receptor tyrosine phosphatase. Cell 73:541–54 105. Palou E, de la Fuente-Garcia MA, Nicolas JM, Vilardell C, Vives J, Gaya A. 1997. CD148, a membrane protein tyrosine phosphatase, is able to induce tyrosine phosphorylation on human lymphocytes. Immunol. Lett. 57:101–3 106. Baker JE, Majeti R, Tangye SG, Weiss A. 2001. Protein tyrosine phosphatase CD148-mediated inhibition of T-cell receptor signal transduction is associated with reduced LAT and phospholipase Cgamma1 phosphorylation. Mol. Cell. Biol. 21:2393–403 107. Tangye SG, Wu J, Aversa G, de Vries JE, Lanier LL, Phillips JH. 1998. Negative regulation of human T cell activation by the receptor-type protein tyrosine phosphatase CD148. J. Immunol. 161:3803–7 108. Han S, Williams S, Mustelin T. 2000. Cytoskeletal protein tyrosine phosphatase PTPH1 reduces T cell antigen receptor signaling. Eur. J. Immunol. 30:1318–25 109. Saxena M, Williams S, Brockdorff J, Gilman J, Mustelin T. 1999. Inhibition of T cell signaling by mitogen-activated protein kinase-targeted hematopoietic tyrosine phosphatase (HePTP). J. Biol. Chem. 274:11,693–700 110. Saxena M, Williams S, Gilman J, Mustelin T. 1998. Negative regulation of T cell antigen receptor signal transduction by hematopoietic tyrosine phosphatase (HePTP). J. Biol. Chem. 273:15,340–44 111. Ward Y, Gupta S, Jensen P, Wartmann M, Davis RJ, Kelly K. 1994. Control of MAP kinase activation by the mitogeninduced threonine/tyrosine phosphatase PAC1. Nature 367:651–54 112. Krystal G. 2000. Lipid phosphatases in

14 Feb 2002

12:31

AR

AR152-21.tex

AR152-21.SGM

LaTeX2e(2001/05/10)

P1: GJC

INHIBITORY SIGNALING

113.

Annu. Rev. Immunol. 2002.20:669-707. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

114.

115.

116.

117.

118.

119.

the immune system. Semin. Immunol. 12:397–403 Rohrschneider LR, Fuller JF, Wolf I, Liu Y, Lucas DM. 2000. Structure, function, and biology of SHIP proteins. Genes Dev. 14:505–20 Damen JE, Liu L, Rosten P, Humphries RK, Jefferson AB, Majerus PW, Krystal G. 1996. The 145-kDa protein induced to associate with Shc by multiple cytokines is an inositol tetraphosphate and phosphatidylinositol 3, 4, 5-triphosphate 5-phosphatase. Proc. Natl. Acad. Sci. USA 93:1689–93 Lioubin MN, Algate PA, Tsai S, Carlberg K, AebersoldA, Rohrschneider LR.1996. p150Ship, a signal transduction molecule with inositol polyphosphate-5-phosphatase activity. Genes Dev. 10:1084–95 Helgason CD, Damen JE, Rosten P, Grewal R, Sorensen P, Chappel SM, Borowski A, Jirik F, Krystal G, Humphries RK. 1998. Targeted disruption of SHIP leads to hemopoietic perturbations, lung pathology, and a shortened life span. Genes Dev. 12:1610–20 Liu Q, Oliveira DSA, Mariathasan S, Bouchard D, Jones J, Sarao R, Kozieradzki I, Ohashi PS, Penninger JM, Dumont DJ. 1998. The inositol polyphosphate 5phosphatase ship is a crucial negative regulator of B cell antigen receptor signaling. J. Exp. Med. 188:1333–42 Brauweiler A, Tamir I, Dal Porto J, Benschop RJ, Helgason CD, Humphries RK, Freed JH, Cambier JC. 2000. Differential regulation of B cell development, activation, and death by the src homology 2 domain-containing 50 inositol phosphatase (SHIP). J. Exp. Med. 191:1545–54 Helgason CD, Kalberer CP, Damen JE, Chappel SM, Pineault N, Krystal G, Humphries RK. 2000. A dual role for Src homology 2 domain-containing inositol5-phosphatase (SHIP) in immunity: aberrant development and enhanced function of b lymphocytes in ship −/− mice. J. Exp. Med. 191:781–94

703

120. Kim CH, Hangoc G, Cooper S, Helgason CD, Yew S, Humphries RK, Krystal G, Broxmeyer HE. 1999. Altered responsiveness to chemokines due to targeted disruption of SHIP. J. Clin. Invest. 104:1751– 59 121. Aman MJ, Lamkin TD, Okada H, Kurosaki T, Ravichandran KS. 1998. The inositol phosphatase SHIP inhibits Akt/PKB activation in B cells. J. Biol. Chem. 273:33922–28 122. Huber M, Helgason CD, Scheid MP, Duronio V, Humphries RK, Krystal G. 1998. Targeted disruption of SHIP leads to Steel factor-induced degranulation of mast cells. EMBO J. 17:7311–19 123. Huber M, Helgason CD, Damen JE, Liu L, Humphries RK, Krystal G. 1998. The src homology 2-containing inositol phosphatase (SHIP) is the gatekeeper of mast cell degranulation. Proc. Natl. Acad. Sci. USA 95:11330–35 124. Okada H, Bolland S, Hashimoto A, Kurosaki M, Kabuyama Y, Iino M, Ravetch JV, Kurosaki T. 1998. Role of the inositol phosphatase SHIP in B cell receptorinduced Ca2+ oscillatory response. J. Immunol. 161:5129–32 125. Hashimoto A, Hirose K, Okada H, Kurosaki T, Iino M. 1999. Inhibitory modulation of B cell receptor-mediated Ca2+ mobilization by Src homology 2 domain-containing inositol 50 -phosphatase (SHIP). J. Biol. Chem. 274:11,203–8 126. Liu Q, Sasaki T, Kozieradzki I, Wakeham A, Itie A, Dumont DJ, Penninger JM. 1999. SHIP is a negative regulator of growth factor receptor-mediated PKB/ Akt activation and myeloid cell survival. Genes Dev. 13:786–91 127. Cox D, Dale BM, Kashiwada M, Helgason CD, Greenberg S. 2001. A regulatory role for Src homology 2 domaincontaining inositol 50 -phosphatase (SHIP) in phagocytosis mediated by Fc gamma receptors and complement receptor 3 (alpha(M)beta(2); CD11b/CD18). J. Exp. Med. 193:61–71

4 Feb 2002

11:58

Annu. Rev. Immunol. 2002.20:669-707. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

704

AR

AR152-21.tex

VEILLETTE

¥

AR152-21.SGM

LATOUR

¥

LaTeX2e(2001/05/10)

P1: GJC

DAVIDSON

128. Latour S, Gish G, Helgason CD, Humphries RK, Pawson T, Veillette A. 2001. Regulation of SLAM-mediated signal transduction by SAP, the X-linked lymphoproliferative gene product. Nat. Immunol. 2:681–90 129. Mikhalap SV, Shlapatska LM, Berdova AG, Law CL, Clark EA, Sidorenko SP. 1999. CDw150 associates with srchomology 2-containing inositol phosphatase and modulates CD95-mediated apoptosis. J. Immunol. 162:5719–27 130. Grimm J, Sachs M, Britsch S, Di Cesare S, Schwarz-Romond T, Alitalo K, Birchmeier W. 2001. Novel p62dok family members, dok-4 and dok-5, are substrates of the c-Ret receptor tyrosine kinase and mediate neuronal differentiation. J. Cell. Biol. 154:345–54 131. Suzu S, Tanaka-Douzono M, Nomaguchi K, Yamada M, Hayasawa H, Kimura F, Motoyoshi K. 2000. p56(dok-2) as a cytokine-inducible inhibitor of cell proliferation and signal transduction. EMBO J. 19:5114–22 132. Tamir I, Stolpa JC, Helgason CD, Nakamura K, Bruhns P, Daeron M, Cambier JC. 2000. The RasGAP-binding protein p62dok is a mediator of inhibitory FcgammaRIIB signals in B cells. Immunity 12:347–58 133. Berg KL, Siminovitch KA, Stanley ER. 1999. SHP-1 regulation of p62(DOK) tyrosine phosphorylation in macrophages. J. Biol. Chem. 274:35855–65 134. Cong F, Yuan B, Goff SP. 1999. Characterization of a novel member of the DOK family that binds and modulates Abl signaling. Mol. Cell. Biol. 19:8314–25 135. Jones N, Dumont DJ. 1998. The Tek/Tie2 receptor signals through a novel Dokrelated docking protein, Dok-R. Oncogene 17:1097–108 136. Nelms K, Snow AL, Hu-Li J, Paul WE. 1998. FRIP, a hematopoietic cellspecific rasGAP-interacting protein phosphorylated in response to cytokine stimulation. Immunity 9:13–24

137. Yamanashi Y, Tamura T, Kanamori T, Yamane H, Nariuchi H, Yamamoto T, Baltimore D. 2000. Role of the rasGAPassociated docking protein p62(dok) in negative regulation of B cell receptormediated signaling. Genes Dev. 14:11–16 138. Jones N, Dumont DJ. 1999. Recruitment of dok-R to the EGF receptor through its PTB domain is required for attenuation of erk MAP kinase activation. Curr. Biol. 9:1057–60 139. Di Cristofano A, Niki M, Zhao M, Karnell FG, Clarkson B, Pear WS, Van Aelst L, Pandolfi PP. 2001. p62(dok), a negative regulator of Ras and mitogen-activated protein kinase (MAPK) activity, opposes leukemogenesis by p210(bcr-abl). J. Exp. Med. 194:275–84 140. Yoshida K, Yamashita Y, Miyazato A, Ohya K, Kitanaka A, Ikeda U, Shimada K, Yamanaka T, Ozawa K, Mano H. 2000. Mediation by the protein-tyrosine kinase Tec of signaling between the B cell antigen receptor and Dok-1. J. Biol. Chem. 275:24945–52 141. Pelicci G, Lanfrancone L, Grignani F, McGlade J, Cavallo F, Forni G, Nicoletti I, Pawson T, Pelicci PG. 1992. A novel transforming protein (SHC) with an SH2 domain is implicated in mitogenic signal transduction. Cell 70:93–104 142. Ingham RJ, Okada H, Dang-Lawson M, Dinglasan J, van Der G, Kurosaki T, Gold MR. 1999. Tyrosine phosphorylation of shc in response to B cell antigen receptor engagement depends on the SHIP inositol phosphatase. J. Immunol. 163:5891–95 143. Galandrini R, Tassi I, Morrone S, Lanfrancone L, Pelicci P, Piccoli M, Frati L, Santoni A. 2001. The adaptor protein shc is involved in the negative regulation of NK cell-mediated cytotoxicity. Eur. J. Immunol. 31:2016–25 144. Pesesse X, Deleu S, De Smedt F, Drayer L, Erneux C. 1997. Identification of a second SH2-domain-containing protein closely related to the phosphatidylinositol polyphosphate 5-phosphatase SHIP.

4 Feb 2002

11:58

AR

AR152-21.tex

AR152-21.SGM

LaTeX2e(2001/05/10)

P1: GJC

INHIBITORY SIGNALING

145.

Annu. Rev. Immunol. 2002.20:669-707. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

146.

147.

148.

149.

150.

151.

152.

153.

Biochem. Biophys. Res. Commun. 239: 697–700 Brauweiler A, Tamir I, Marschner S, Helgason CD, Cambier JC. 2001. Partially distinct molecular mechanisms mediate inhibitory FcgammaRIIB signaling in resting and activated B cells. J. Immunol. 167:204–11 Clement S, Krause U, Desmedt F, Tanti JF, Behrends J, Pesesse X, Sasaki T, Penninger J, Doherty M, Malaisse W, Dumont JE, Marchand-Brustel Y, Erneux C, Hue L, Schurmans S. 2001. The lipid phosphatase SHIP2 controls insulin sensitivity. Nature 409:92–97 Simpson L, Parsons R. 2001. PTEN: life as a tumor suppressor. Exp. Cell. Res. 264:29–41 Di Cristofano A, Pandolfi PP. 2000. The multiple roles of PTEN in tumor suppression. Cell 100:387–90 Vazquez F, Sellers WR. 2000. The PTEN tumor suppressor protein: an antagonist of phosphoinositide 3-kinase signaling. Biochim. Biophys. Acta 1470:M21–35 Suzuki A, de la Pompa JL, Stambolic V, Elia AJ, Sasaki T, del BB I, Ho A, Wakeham A, Itie A, Khoo W, Fukumoto M, Mak TW. 1998. High cancer susceptibility and embryonic lethality associated with mutation of the PTEN tumor suppressor gene in mice. Curr. Biol. 8:1169–78 Suzuki A, Yamaguchi MT, Ohteki T, Sasaki T, Kaisho T, Kimura Y, Yoshida R, Wakeham A, Higuchi T, Fukumoto M, Tsubata T, Ohashi PS, Koyasu S, Penninger JM, Nakano T, Mak TW. 2001. T cellspecific loss of Pten leads to defects in central and peripheral tolerance. Immunity. 14:523–34 Wang X, Gjorloff-Wingren A, Saxena M, Pathan N, Reed JC, Mustelin T. 2000. The tumor suppressor PTEN regulates T cell survival and antigen receptor signaling by acting as a phosphatidylinositol 3-phosphatase. J. Immunol. 164:1934– 39 Shan X, Czar MJ, Bunnell SC, Liu P, Liu

154.

155.

156.

157.

158.

159.

160.

161.

705

Y, Schwartzberg PL, Wange RL. 2000. Deficiency of PTEN in Jurkat T cells causes constitutive localization of Itk to the plasma membrane and hyperresponsiveness to CD3 stimulation. Mol. Cell. Biol. 20:6945–57 Chan VW, Meng F, Soriano P, DeFranco AL, Lowell CA. 1997. Characterization of the B lymphocyte populations in Lyndeficient mice and the role of Lyn in signal initiation and down-regulation. Immunity 7:69–81 Nishizumi H, Taniuchi I, Yamanashi Y, Kitamura D, Ilic D, Mori S, Watanabe T, Yamamoto T. 1995. Impaired proliferation of peripheral B cells and indication of autoimmune disease in lyn-deficient mice. Immunity 3:549–60 Hibbs ML, Tarlinton DM, Armes J, Grail D, Hodgson G, Maglitto R, Stacker SA, Dunn AR. 1995. Multiple defects in the immune system of Lyn-deficient mice, culminating in autoimmune disease. Cell 83:301–11 Chan VW, Lowell CA, DeFranco AL. 1998. Defective negative regulation of antigen receptor signaling in Lyn-deficient B lymphocytes. Curr. Biol. 8:545– 53 Nishizumi H, Horikawa K, MlinaricRascan I, Yamamoto T. 1998. A doubleedged kinase Lyn: a positive and negative regulator for antigen receptor-mediated signals. J. Exp. Med. 187:1343–48 Smith KG, Tarlinton DM, Doody GM, Hibbs ML, Fearon DT. 1998. Inhibition of the B cell by CD22: a requirement for Lyn. J. Exp. Med. 187:807–11 Gresham HD, Dale BM, Potter JW, Chang PW, Vines CM, Lowell CA, Lagenaur CF, Willman CL. 2000. Negative regulation of phagocytosis in murine macrophages by the Src kinase family member, Fgr. J. Exp. Med. 191:515–28 Thien CB, Langdon WY. 2001. Cbl: many adaptations to regulate protein tyrosine kinases. Nat. Rev. Mol. Cell. Biol. 2:294– 307

14 Feb 2002

12:31

Annu. Rev. Immunol. 2002.20:669-707. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

706

AR

AR152-21.tex

VEILLETTE

¥

AR152-21.SGM

LATOUR

¥

LaTeX2e(2001/05/10)

P1: GJC

DAVIDSON

162. Rudd CE, Schneider H. 2000. Lymphocyte signaling: Cbl sets the threshold for autoimmunity. Curr. Biol. 10:R344–47 163. Lupher ML Jr, Rao N, Eck MJ, Band H. 1999. The Cbl protooncoprotein: a negative regulator of immune receptor signal transduction. Immunol. Today 20:375–82 164. Langdon WY, Hyland CD, Grumont RJ, Morse HC III. 1989. The c-cbl protooncogene is preferentially expressed in thymus and testis tissue and encodes a nuclear protein. J. Virol. 63:5420–24 165. Langdon WY, Hartley JW, Klinken SP, Ruscetti SK, Morse HC III. 1989. v-cbl, an oncogene from a dual-recombinant murine retrovirus that induces early B-lineage lymphomas. Proc. Natl. Acad. Sci. USA 86:1168–72 166. Blake TJ, Langdon WY. 1992. A rearrangement of the c-cbl proto-oncogene in HUT78 T-lymphoma cells results in a truncated protein. Oncogene 7:757–62 167. Blake TJ, Shapiro M, Morse HC III, Langdon WY. 1991. The sequences of the human and mouse c-cbl proto-oncogenes show v-cbl was generated by a large truncation encompassing a proline-rich domain and a leucine zipper-like motif. Oncogene 6:653–57 168. Lupher ML Jr, Rao N, Lill NL, Andoniou CE, Miyake S, Clark EA, Druker B, Band H. 1998. Cbl-mediated negative regulation of the Syk tyrosine kinase. A critical role for Cbl phosphotyrosine-binding domain binding to Syk phosphotyrosine 323. J. Biol. Chem. 273:35,273–81 169. Rao N, Lupher ML Jr, Ota S, Reedquist KA, Druker BJ, Band H. 2000. The linker phosphorylation site Tyr292 mediates the negative regulatory effect of Cbl on ZAP70 in T cells. J. Immunol. 164:4616–26 170. Yankee TM, Keshvara LM, Sawasdikosol S, Harrison ML, Geahlen RL. 1999. Inhibition of signaling through the B cell antigen receptor by the protooncogene product, c-Cbl, requires Syk tyrosine 317 and the c-Cbl phosphotyrosine-binding domain. J. Immunol. 163:5827–35

171. Ota S, Hazeki K, Rao N, Lupher ML Jr, Andoniou CE, Druker B, Band H. 2000. The RING finger domain of Cbl is essential for negative regulation of the Syk tyrosine kinase. J. Biol. Chem. 275:414– 22 172. Andoniou CE, Lill NL, Thien CB, Lupher ML Jr, Ota S, Bowtell DD, Scaife RM, Langdon WY, Band H. 2000. The Cbl proto-oncogene product negatively regulates the Src-family tyrosine kinase Fyn by enhancing its degradation. Mol. Cell. Biol. 20:851–67 173. Thien CB, Bowtell DD, Langdon WY. 1999. Perturbed regulation of ZAP-70 and sustained tyrosine phosphorylation of LAT and SLP-76 in c-Cbl-deficient thymocytes. J. Immunol. 162:7133–39 174. Naramura M, Kole HK, Hu RJ, Gu H. 1998. Altered thymic positive selection and intracellular signals in Cbl-deficient mice. Proc. Natl. Acad. Sci. USA 95: 15547–52 175. Murphy MA, Schnall RG, Venter DJ, Barnett L, Bertoncello I, Thien CB, Langdon WY, Bowtell DD. 1998. Tissue hyperplasia and enhanced T-cell signalling via ZAP-70 in c-Cbl-deficient mice. Mol. Cell. Biol. 18:4872–82 176. Wang HY, Altman Y, Fang D, Elly C, Dai Y, Shao Y, Liu YC. 2001. Cbl promotes ubiquitination of the T cell receptor zeta through an adaptor function of Zap-70. J. Biol. Chem. 276:26,004–11 177. Yasuda T, Maeda A, Kurosaki M, Tezuka T, Hironaka K, Yamamoto T, Kurosaki T. 2000. Cbl suppresses B cell receptor-mediated phospholipase C (PLC)gamma2 activation by regulating B cell linker protein-PLC-gamma2 binding. J. Exp. Med. 191:641–50 178. Chiang YJ, Kole HK, Brown K, Naramura M, Fukuhara S, Hu RJ, Jang IK, Gutkind JS, Shevach E, Gu H. 2000. Cbl-b regulates the CD28 dependence of T-cell activation. Nature 403:216–20 179. Bachmaier K, Krawczyk C, Kozieradzki I, Kong YY, Sasaki T, Oliveira-dos-Santos

4 Feb 2002

11:58

AR

AR152-21.tex

AR152-21.SGM

LaTeX2e(2001/05/10)

P1: GJC

INHIBITORY SIGNALING

Annu. Rev. Immunol. 2002.20:669-707. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

A, Mariathasan S, Bouchard D, Wakeham A, Itie A, Le J, Ohashi PS, Sarosi I, Nishina H, Lipkowitz S, Penninger JM. 2000. Negative regulation of lymphocyte activation and autoimmunity by the molecular adaptor Cbl-b. Nature 403:211– 16 180. Krawczyk C, Bachmaier K, Sasaki T, Jones GR, Snapper BS, Bouchard D,

707

Kozieradzki I, Ohashi SP, Alt WF, Penninger MJ. 2000. Cbl-b is a negative regulator of receptor clustering and raft aggregation in T cells. Immunity 13:463–73 181. Fang D, Wang HY, Fang N, Altman Y, Elly C, Liu YC. 2001. Cbl-b, a RING-type E3 ubiquitin ligase, targets phosphatidylinositol 3-kinase for ubiquitination in T cells. J. Biol. Chem. 276:4872–78

15:15 AR AR152-21-COLOR.tex AR152-21-COLOR.SGM

Figure 1 Conformation of activated and inactive Src kinases. Schematic representations of the structure of a Src kinase (Lck), either activated or inactive. In the inactive form, intramolecular interactions exist between the SH2 domain and the phosphorylated carboxylterminus, and between the SH3 domain and the linker region.

Annu. Rev. Immunol. 2002.20:669-707. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

14 Feb 2002 LaTeX2e(2002/01/18) P1: GDL

15:15 AR AR152-21-COLOR.tex AR152-21-COLOR.SGM

Figure 2 Model for T-cell inhibition by the PAG-Csk complex. See text for details.

Annu. Rev. Immunol. 2002.20:669-707. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

14 Feb 2002 LaTeX2e(2002/01/18) P1: GDL

15:15 AR AR152-21-COLOR.tex AR152-21-COLOR.SGM

Figure 5 Differential effects of PTP-PEST and SHP-1 on BCR signaling. See text for details.

Annu. Rev. Immunol. 2002.20:669-707. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

14 Feb 2002 LaTeX2e(2002/01/18) P1: GDL

15:15 AR AR152-21-COLOR.tex AR152-21-COLOR.SGM LaTeX2e(2002/01/18)

Figure 6 Model for negative regulation of BCR signaling by a Dok-related adaptor. See text for details.

Annu. Rev. Immunol. 2002.20:669-707. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

14 Feb 2002 P1: GDL

P1: FRK

February 1, 2002

12:47

Annual Reviews

AR152-FM

Annual Review of Immunology Volume 20, 2002

Annu. Rev. Immunol. 2002.20:669-707. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

CONTENTS FRONTISPIECE, Charles A. Janeway, Jr. A TRIP THROUGH MY LIFE WITH AN IMMUNOLOGICAL THEME, Charles A. Janeway, Jr.

xiv 1

THE B7 FAMILY OF LIGANDS AND ITS RECEPTORS: NEW PATHWAYS FOR COSTIMULATION AND INHIBITION OF IMMUNE RESPONSES, Beatriz M. Carreno and Mary Collins

29

MAP KINASES IN THE IMMUNE RESPONSE, Chen Dong, Roger J. Davis, and Richard A. Flavell

PROSPECTS FOR VACCINE PROTECTION AGAINST HIV-1 INFECTION AND AIDS, Norman L. Letvin, Dan H. Barouch, and David C. Montefiori T CELL RESPONSE IN EXPERIMENTAL AUTOIMMUNE ENCEPHALOMYELITIS (EAE): ROLE OF SELF AND CROSS-REACTIVE ANTIGENS IN SHAPING, TUNING, AND REGULATING THE AUTOPATHOGENIC T CELL REPERTOIRE, Vijay K. Kuchroo, Ana C. Anderson, Hanspeter Waldner, Markus Munder, Estelle Bettelli, and Lindsay B. Nicholson

55 73

101

NEUROENDOCRINE REGULATION OF IMMUNITY, Jeanette I. Webster, Leonardo Tonelli, and Esther M. Sternberg

125

MOLECULAR MECHANISM OF CLASS SWITCH RECOMBINATION: LINKAGE WITH SOMATIC HYPERMUTATION, Tasuku Honjo, Kazuo Kinoshita, and Masamichi Muramatsu

165

INNATE IMMUNE RECOGNITION, Charles A. Janeway, Jr. and Ruslan Medzhitov

KIR: DIVERSE, RAPIDLY EVOLVING RECEPTORS OF INNATE AND ADAPTIVE IMMUNITY, Carlos Vilches and Peter Parham ORIGINS AND FUNCTIONS OF B-1 CELLS WITH NOTES ON THE ROLE OF CD5, Robert Berland and Henry H. Wortis E PROTEIN FUNCTION IN LYMPHOCYTE DEVELOPMENT, Melanie W. Quong, William J. Romanow, and Cornelis Murre

197 217 253 301

LYMPHOCYTE-MEDIATED CYTOTOXICITY, John H. Russell and Timothy J. Ley

SIGNAL TRANSDUCTION MEDIATED BY THE T CELL ANTIGEN x

323

P1: FRK

Annu. Rev. Immunol. 2002.20:669-707. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

February 1, 2002

12:47

Annual Reviews

AR152-FM

CONTENTS

xi

RECEPTOR: THE ROLE OF ADAPTER PROTEINS, Lawrence E. Samelson INTERACTION OF HEAT SHOCK PROTEINS WITH PEPTIDES AND ANTIGEN PRESENTING CELLS: CHAPERONING OF THE INNATE AND ADAPTIVE IMMUNE RESPONSES, Pramod Srivastava CHROMATIN STRUCTURE AND GENE REGULATION IN THE IMMUNE SYSTEM, Stephen T. Smale and Amanda G. Fisher PRODUCING NATURE’S GENE-CHIPS: THE GENERATION OF PEPTIDES FOR DISPLAY BY MHC CLASS I MOLECULES, Nilabh Shastri, Susan

371

Schwab, and Thomas Serwold

395 427

463

THE IMMUNOLOGY OF MUCOSAL MODELS OF INFLAMMATION, Warren Strober, Ivan J. Fuss, and Richard S. Blumberg T CELL MEMORY, Jonathan Sprent and Charles D. Surh

GENETIC DISSECTION OF IMMUNITY TO MYCOBACTERIA: THE HUMAN MODEL, Jean-Laurent Casanova and Laurent Abel ANTIGEN PRESENTATION AND T CELL STIMULATION BY DENDRITIC CELLS, Pierre Guermonprez, Jenny Valladeau, Laurence Zitvogel, Clotilde Th´ery, and Sebastian Amigorena

495 551 581

621

NEGATIVE REGULATION OF IMMUNORECEPTOR SIGNALING, Andr´e Veillette, Sylvain Latour, and Dominique Davidson

669

CPG MOTIFS IN BACTERIAL DNA AND THEIR IMMUNE EFFECTS, Arthur M. Krieg

PROTEIN KINASE Cθ

709 IN

T CELL ACTIVATION, Noah Isakov and Amnon

Altman

RANK-L AND RANK: T CELLS, BONE LOSS, AND MAMMALIAN EVOLUTION, Lars E. Theill, William J. Boyle, and Josef M. Penninger PHAGOCYTOSIS OF MICROBES: COMPLEXITY IN ACTION, David M. Underhill and Adrian Ozinsky

761 795 825

STRUCTURE AND FUNCTION OF NATURAL KILLER CELL RECEPTORS: MULTIPLE MOLECULAR SOLUTIONS TO SELF, NONSELF DISCRIMINATION, Kannan Natarajan, Nazzareno Dimasi, Jian Wang, Roy A. Mariuzza, and David H. Margulies

853

INDEXES Subject Index Cumulative Index of Contributing Authors, Volumes 1–20 Cumulative Index of Chapter Titles, Volumes 1–20

ERRATA An online log of corrections to Annual Review of Immunology chapters (if any, 1997 to the present) may be found at http://immunol.annualreviews.org/

887 915 925

8 Feb 2002

10:8

AR

AR152-22.tex

AR152-22.SGM

LaTeX2e(2001/05/10) P1: GJC 10.1146/annurev.immunol.20.100301.064842

Annu. Rev. Immunol. 2002. 20:709–760 DOI: 10.1146/annurev.immunol.20.100301.064842 c 2002 by Annual Reviews. All rights reserved Copyright °

CPG MOTIFS IN BACTERIAL DNA ∗ AND THEIR IMMUNE EFFECTS Annu. Rev. Immunol. 2002.20:709-760. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

Arthur M. Krieg Department of Veterans Affairs Medical Center, Iowa City, Iowa 52246, and Department of Internal Medicine, University of Iowa College of Medicine, Iowa City, Iowa 52242, and Coley Pharmaceutical Group, 93 Worcester Street, Suite 101, Wellesley, Massachussetts 02481; e-mail: [email protected]

■ Abstract Unmethylated CpG motifs are prevalent in bacterial but not vertebrate genomic DNAs. Oligodeoxynucleotides (ODN) containing CpG motifs activate host defense mechanisms leading to innate and acquired immune responses. The recognition of CpG motifs requires Toll-like receptor (TLR) 9, which triggers alterations in cellular redox balance and the induction of cell signaling pathways including the mitogen activated protein kinases (MAPKs) and NFκB. Cells that express TLR-9, which include plasmacytoid dendritic cells (PDCs) and B cells, produce Th1-like proinflammatory cytokines, interferons, and chemokines. Certain CpG motifs (CpG-A) are especially potent at activating NK cells and inducing IFN-α production by PDCs, while other motifs (CpG-B) are especially potent B cell activators. CpG-induced activation of innate immunity protects against lethal challenge with a wide variety of pathogens, and has therapeutic activity in murine models of cancer and allergy. CpG ODN also enhance the development of acquired immune responses for prophylactic and therapeutic vaccination.

INTRODUCTION Innate immune cells such as dendritic cells (DC) must be activated in order to trigger the generation of optimal adaptive immune responses. These cells of the innate immune system lack the highly specific antigen receptors of T and B cells; instead, they rely on a set of pattern recognition receptors (PRRs), which have a general ability to detect certain molecular structures present in pathogens but not ∗

Abbreviations: APC, antigen presenting cell; BCR, B cell receptor; bDNA, bacterial DNA; CFA, complete Freund’s adjuvant; CpG, cytosine linked to a guanine by a phosphate bond; CpG-A ODN, an ODN containing a CpG motif with bases linked by phosphodiester bonds; CpG-B ODN, an ODN containing a CpG motif with bases linked by phosphorothioate bonds; Id, idiotype; IFN, interferon; LPS, lipopolysaccharide; MAPK, mitogen-activated protein kinase; ODN, oligodeoxynucleotide; PO, phosphodiester; PS, phosphorothioate; TCR, T cell receptor. 0732-0582/02/0407-0709$14.00

709

1 Feb 2002

9:54

Annu. Rev. Immunol. 2002.20:709-760. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

710

AR

AR152-22.tex

AR152-22.SGM

LaTeX2e(2001/05/10)

P1: GJC

KRIEG

in self-tissues (1, 2). Thus, the immune system appears to use the presence of these molecular structures as a danger signal that indicates the presence of infection and activates appropriate defense pathways. Manipulation of PRR-activated pathways may be useful both for the purposes of activating therapeutic responses in cancer or some infectious diseases and also for preventing undesirable immune activation in other clinical settings. The purpose of this review is to consider a recently described PRR ligand, unmethylated CpG dinucleotides in particular base contexts. These CpG motifs are prevalent in bacterial and many viral DNAs but are heavily suppressed and methylated in vertebrate genomes (1–5). Recent studies show that the immune system responds to CpG motifs by activating potent Th1-like immune responses that can be harnessed for immune therapy of cancer, allergy, and infectious diseases.

HISTORY Bacterial Extracts and DNA More than a century has passed since William Coley introduced the deliberate use of bacterial extracts for the successful treatment of cancer (6). Since then, extracts of the attenuated mycobacteria bacillus Calmette Guerin (BCG) have become standard therapy for human bladder cancer (7). It is surprising that the active component of BCG for activating natural killer (NK)cells and inducing tumor regression in mice was the DNA (8). Purified BCG DNA induces NK cell activity and the production of type 1 and type 2 interferons in vitro (9), and it gave some encouraging responses in human clinical trials in Japan (10). By cloning mycobacterial genes and synthesizing constituent oligodeoxynucleotides (ODN), these investigators concluded that certain self-complementary palindromes in these ODN were responsible for the immune stimulatory effects (11). The active palindromes contained at least one CpG dinucleotide and were more common in the genomes of bacteria compared to humans (12). Methylation of the CpGs was reported to have no influence on the immune stimulatory activities of the DNA, which were thought to depend on their secondary or tertiary structure (12). The self-complementary polynucleotide, poly-(dG,dC), also triggered NK cell activation (13). Pisetsky and colleagues independently reported that purified bacterial DNA (bDNA) and poly-(dC,dG) induced murine B cell proliferation and immunoglobulin secretion, but vertebrate DNA did not (14). The mitogenicity of poly-(dC,dG) was abolished by cytosine methylation (15), but these investigators also did not associate CpG methylation with different immune activities of bacterial and vertebrate DNAs. Instead, they suggested that the methylation disrupted unique higher ordered structures of the bDNA molecules. DNA containing runs of consecutive guanines, termed poly G sequences or G-quartets, can induce B cell proliferation (16). bDNA is immune stimulatory when added to cell cultures or injected in mice; however, if the DNA is introduced directly into the cell cytoplasm, then vertebrate

1 Feb 2002

9:54

AR

AR152-22.tex

AR152-22.SGM

LaTeX2e(2001/05/10)

P1: GJC

CPG MOTIFS

711

DNA or fragments of double-stranded DNA or RNA as short as 25 bp trigger nonimmune cells to induce or activate STAT1, STAT3, NFκB, and mitogen-activated protein kinases, and to express MHC and costimulatory molecules in a sequenceindependent manner (17). In contrast to bDNA, which is active in either double- or single-stranded form, vertebrate DNA is only active in the double-stranded form (17). Introduction of host DNA into the cytoplasm of DC induces them to mature with enhanced functional activity (18).

Annu. Rev. Immunol. 2002.20:709-760. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

Antisense Oligodeoxynucleotides Antisense ODN are designed to be complementary to a specific target gene and were initially proposed as an example of rational drug design. To retard degradation, they are generally made with a modified nuclease-resistant backbone— typically a phosphorothioate (PS) backbone in which one of the nonbridging oxygen atoms at each of the phosphodiester (PO) linkages is replaced with a sulfur. Early in the development of antisense technology, several investigators using antisense ODN reported unexpected stimulation of lymphocyte proliferation and variable effects on immunoglobulin production. An antisense ODN to an immunoglobulin sequence unexpectedly increased the expression of the target gene and induced B cell proliferation but inhibited antibody secretion (19). Some but not all control ODNs in which various bases were switched also had immune stimulatory activities. An antisense ODN against the rev gene of the human immunodeficiency virus (HIV) caused a profound degree of B cell proliferation and massive splenomegaly in vivo in mice (20). In contrast to the inhibitory effects observed in the previous study, the anti-rev ODN induced production of immunoglobulin. An antisense ODN against herpes simplex virus (HSV) also induced B cell proliferation, but control ODN did not (21). A sense ODN against the nuclear factor κB (NFκB) p65 subunit actually triggered the rapid activation of NFκB binding activity and induced B cell proliferation, immunoglobulin secretion, and in vivo splenomegaly, but a complementary antisense ODN did not (22). There was no apparent common motif that could be discerned which may tie these different observations together (22).

The Identification of the CpG Motif I was also attempting to perform antisense experiments when I also observed the immune stimulatory effects of certain ODN. However, in my own case, the initial results appeared to be consistent with an antisense mechanism of action and the observed biological effects were therefore interpreted in this manner. We had observed that a particular murine endogenous retroviral sequence, mink cell focus-forming (MCF), was highly expressed in mice genetically predisposed to the development of systemic lupus erythematosus but not in control mice (23). Antisense ODN appeared to be an ideal technology with which to investigate the possibility that some protein product encoded by this retrovirus might possibly have an immune regulatory activity. For my first experiments to test this hypothesis, I

1 Feb 2002

9:54

Annu. Rev. Immunol. 2002.20:709-760. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

712

AR

AR152-22.tex

AR152-22.SGM

LaTeX2e(2001/05/10)

P1: GJC

KRIEG

synthesized two partially overlapping antisense ODN against the MCF retrovirus and two control ODNs (24). This class of endogenous retrovirus contains an immunosuppressive domain in the envelope region, and so it seemed possible that inhibition of expression of this gene could release the lymphocytes from possible suppressive effects, thereby leading to proliferation. The first experiment with these ODNs appeared to be a complete success in that both of the antisense ODNs caused B cell proliferation and immunoglobulin secretion, but both of the controls were nonstimulatory (24). Further studies confirmed and extended these findings (25, 26). Although our results with these immune stimulatory oligos were initially interpreted in the context of an antisense mechanism of action, with further studies it became clear that this could not explain all of the observed activities. Several control ODNs synthesized for other experiments showed immune stimulatory effects with activation of B cell proliferation and immunoglobulin secretion, which was very similar to that induced by the initial antisense ODN. Although we initially considered the possibility that these immune stimulatory control ODN may in fact be antisense to some previously unidentified gene, they showed no homology to any known DNA sequence and did not hybridize to Northern blots of B cell mRNA. Moreover, certain additional control ODN had similar effects as opposed to others, which did not. It was therefore clear that this was some type of unexpected and sequence-specific but nonantisense activity. The magnitude of the immune stimulatory effect was startling; the strongest ODN were more effective B cell mitogens than gold standards such as endotoxin. Initially it was unclear whether there was any common mechanism between the immune stimulatory effects triggered by these ODN and those reported in the previous studies of antisense ODN reviewed in the preceding section. Furthermore, very few of the ODN contained any palindromes, so this appeared to be a different type of immune activity than that reported by Tokunaga and colleagues for the activation of NK cells. In order to identify the DNA structure responsible for these effects, a series of structure/function analyses were therefore undertaken. Initial studies suggested that some sort of secondary structure such as a stem loop could explain at least some of the immune stimulatory activities caused by the particular control ODN. This raised the possibility that there may be several different structures or sequences capable of mediating lymphocyte activation. However, after synthesizing and testing several hundred ODN, it eventually became clear that a sufficient sequence element for inducing B cell activation was a CpG dinucleotide in particular base contexts. Further experiments confirmed that stem loop structures or other secondary structures were not required for immune stimulation. In addition, with the understanding of the role of flanking bases around the CpGs in determining immune stimulation, it became possible to review with a new eye previous reports of immune stimulatory antisense ODN and palindromes. This comparison revealed that all of the previously reported immune stimulatory sequences also contained CpG dinucleotides, which conformed to the general consensus motif of XCGY, where X is any base but C, and Y is any base but G (4, 27). It is interesting that

1 Feb 2002

9:54

AR

AR152-22.tex

AR152-22.SGM

LaTeX2e(2001/05/10)

P1: GJC

CPG MOTIFS

713

while in general, increasing the number of stimulatory CpG motifs in an ODN increased the activity of the ODN, the addition of a CpG into an end of an ODN or the addition of a CpG in an unfavorable sequence context could actually reduce the degree of the cell activation (4, 28). Elimination of the CpG dinucleotides from ODN abolished their stimulatory activities (4).

Annu. Rev. Immunol. 2002.20:709-760. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

Immune Recognition of CpG Motifs as a Defense Mechanism While the identification of the CpG motif explained the way in which certain ODN caused immune stimulatory effects, it raised a new question of what purpose these effects could possibly serve. Vertebrate and bacterial DNAs differ markedly in their CpG content owing to CpG suppression: CpG dinucleotides in vertebrate genomes occur only about a quarter as frequently as would be predicted if base utilization was random (3). Furthermore, the bases flanking CpGs in vertebrate genomes are not random: The most common base preceding a CpG is a C and the most common base following a CpG is a G (29). As noted above, these types of CpG motifs do not support strong immune stimulation of a CpG. In addition to these differences in CpG content, CpG dinucleotides are not methylated in bDNA but are routinely methylated at the 5 position of about 70% of the cytosines in vertebrate DNAs (3). This structural difference between bacterial and vertebrate DNAs suggested the possibility that immune recognition of unmethylated CpG motifs in bDNA could explain the previously reported effects of bDNA on B and NK cells rather than the previously proposed mechanisms involving palindromes and higher ordered structures. To test this hypothesis, we synthesized immune stimulatory ODN in which the cytosines of the CpG were replaced by 5-methyl cytosine. These ODN lost their immune stimulatory effects, as opposed to ODN, in which other cytosines outside of the CpG motifs were replaced by 5-methyl cytosine, which retained full immune stimulatory capacities (4). Moreover, the immune stimulatory effects of bDNA were also completely abolished by methylation with CpG methylase (4). Extracts of Babesia bovis, like those of many other microbes, are immune stimulatory. These effects result from the DNAbecause they are abolished by treatment of the extracts with nuclease and can be reproduced with purified B. bovis DNA (30). Immune stimulatory DNA is not unique to microbes; Drosophila extracts are also immune stimulatory, owing to the unmethylated CpG motifs in insect DNA (31). A survey of different genomic DNAs has shown immune stimulation by nematode and mollusc DNAs and confirmed that hypomethylation of CpG is required for immune stimulation (32). These data support the concept that immune recognition of CpG motifs triggers protective pathways analogous to those activated by the PRRs that detect endotoxins and other microbial products. If immune recognition of CpG DNA has evolved as an effective defense, then it seems likely that pathogens would evolve counter strategies to block or evade the defense. One potential strategy would be for a pathogen to reduce the level of CpGs in its genome to decrease its immune

1 Feb 2002

9:54

Annu. Rev. Immunol. 2002.20:709-760. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

714

AR

AR152-22.tex

AR152-22.SGM

LaTeX2e(2001/05/10)

P1: GJC

KRIEG

stimulatory effects. In principle, this should be much easier for pathogens with small genomes than for those with large genomes. Indeed, statistical analyses of the genomes of small DNA viruses and retroviruses have revealed that these all have very low CpG content (33–35). Among the organisms with the lowest CpG contents are those that are associated with long-term stable infection, such as JC virus and papova viruses (33). CpG suppression has also been noted in the genomes of several intracellular parasites, such as Plasmodium falciparum and Entamoeba histolytica (33), and several bacterial species, including Mycobacterium jannaschii, Mycobacterium genitalium, and Borrelia burgdorferi, which have 30%–50% of the expected CpG frequency, while some closely related bacteria, such as Mycobacterium pneumoniae, have no CpG suppression (36). Further studies will be required to determine whether these distinct patterns of CpG suppression in different species are broadly associated with different patterns of infection, such as organisms that cause more chronic invasive infections having more marked CpG suppression.

TYPES OF CpG DNA AND ODN Identification of Optimal CpG Motifs In general, CpG DNA stimulates B cells, NK cells, DC, and monocytes/macrophages, regardless of whether the DNA is in the form of genomic bDNA or in the form of synthetic ODN with a nuclease-resistant PS backbone. However, the level of immune stimulatory effects of an ODN depends to a great degree on the precise bases flanking the CpG dinucleotide. Together with the one or two bases on its 50 and 30 sides, the CpG dinucleotide comprises a CpG motif. By examining many possible base combinations, optimal CpG motifs for activating mouse or rabbit immune cells have the general formula, purine-purine-CG-pyrimidine-pyrimidine; however, the best CpG motif was found to be GACGTT (4, 37, 38). For activating human cells, the optimal motif is GTCGTT (39). A recent study suggests that human T and NK cells may respond to different CpG motifs (40), but this requires confirmation and further analysis. The human GTCGTT motif also appears to be optimal for many other vertebrate species, including cow, sheep, cat, dog, goat, horse, pig, and chicken (30, 38). Fish immune cells are activated by CpG DNA but may respond to different CpG motifs than mice (41). The immune stimulatory effects of the ODN are enhanced if the ODN has a TpC dinucleotide on the 50 end and is pyrimidine rich on the 30 side (37, 39, 42). The immune stimulatory properties of a particular ODN or fragment of DNA are also affected by the number and spacing of the CpG motifs, the presence of poly G sequences or other flanking sequences in the ODN, and the ODN backbone (4, 16, 37, 39, 42, 43). Although the most potent ODN usually have two or three CpG motifs, the addition of more than three optimal motifs does not increase activity further. Long DNA sequences are not required for immune activation; sequences as short as six bases show some activity, though perhaps not the full range of effects that can be observed with longer ODN (44). For optimal stimulatory

1 Feb 2002

9:54

AR

AR152-22.tex

AR152-22.SGM

LaTeX2e(2001/05/10)

P1: GJC

CPG MOTIFS

715

activity, the CpG motifs in ODN containing several motifs should not be backto-back but preferably spaced with at least two intervening bases, preferably Ts (37, 39, 42).

Annu. Rev. Immunol. 2002.20:709-760. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

Effects of the ODN Backbone on CpG-Mediated Immune Activation ODN constructed with the normal PO DNA backbone are rapidly degraded inside lymphocytes (45). The susceptibility of PO ODN to degradation not only reduces their ability to drive B cell proliferation but can even result in an artifact in studies using a 3H thymidine incorporation assay to measure B cell proliferation. In such assays PO DNA can be degraded, releasing free thymidine that competes for the 3 H thymidine and causes a false suppression of incorporation (46). Moreover, nuclease activities can be B cell-specific (47) and higher in human than in mouse cells, with the result that PO DNA can appear nonstimulatory (48) unless it is added repeatedly (49). The PS backbone provides an extremely high degree of nuclease resistance and greatly stabilizes the ODN against degradation (50), with an approximate 200fold increase in the level of CpG-induced B cell activation (4). The PS backbone dramatically increases the nonspecific ODN binding to a wide variety of proteins (51, 52). PS ODN bind much more avidly to cell membranes and generally have a much higher degree of cell uptake (45, 53, 54). The PS backbone results in sequence-independent activities including the activation of SP1 transcription factor activity (55), inhibition of smooth muscle cell proliferation and migration (56, 57), inhibition of basic fibroblast growth factor binding to its receptor (58, 59) and angiogenic activity (60), reduction of the sequence-specific binding of transcription factors to their binding sites (61), inhibition of cellular adhesion to extracellular matrix (62), enhancement of LPS-induced tumor necrosis factor production (63), and some degree of nonsequence-specific immune stimulation (64). Finally, the PS backbone enhances the ability of poly G sequences to inhibit CD28 expression and in vivo contact hypersensitivity responses (65). The immune effects of PS ODN are reduced by ODN modification with 20 methoxyethoxy (66). In general, PS ODN are much more potent at activating B cells compared to the same sequence with a PO backbone (4, 45, 67, 68), but there are some differences in the immune stimulatory effects of CpG motifs in PS ODN compared to PO ODN. PS ODN bearing suboptimal CpG motifs are less likely than PO ODN to drive B cell proliferation, especially if the CpG is followed by a G. PS ODN without CG motifs are also frequently observed to drive the proliferation of murine and human B cells, although to a more limited degree than that which occurs with CpG ODN (37, 42, 48, 64). In contrast, PS ODN are generally less active at activating macrophages or NK cells, compared to ODN in which at least part of the backbone is PO (43, 69, 70). The PS backbone modification creates a chiral center at each internucleotide linkage. Therefore, a 20-mer ODN has 19 chiral centers, giving 219 (approximately

1 Feb 2002

9:54

716

AR

AR152-22.tex

AR152-22.SGM

LaTeX2e(2001/05/10)

P1: GJC

KRIEG

100,000) different stereoisomers. In vivo, the S-isomer is highly resistant to nuclease degradation, but the R-isomer is degraded at a more rapid rate than a stereorandom ODN (71). Consistent with their difference in stability, CpG ODN synthesized with an R backbone are less potent at inducing B cell proliferation compared to the S isomer (72).

Annu. Rev. Immunol. 2002.20:709-760. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

Identification of Distinct Classes of Immune Stimulatory CpG ODN; Characteristics of CpG-A ODN As a result of the effects reviewed above, ODN with different backbones and different sequence motifs can induce dramatically different profiles and kinetics of immune activation (73–79). The term CpG DNA should therefore be used with care because not all of the effects described in the literature are seen with all ODN, bacterial genomic DNAs, or plasmids containing CpG motifs. Throughout this review, effort is made to clarify when the effects described are unique to one or another form of CpG DNA. ODN containing PO backbones are particularly effective at activating NK cells and inducing IFN-α production from plasmacytoid DC precursors, for which reason we have termed them CpG-A ODN (43, 68, 78, 80–82). The highest degrees of NK cell activation and IFN-α production occur with ODN in which the 50 and 30 ends are PS-modified, and the center portion is PO (43). These chimeric ODN have a high degree of nuclease resistance provided by the PS-modified ends, yet they also retain potent NK-stimulating effects because the CpG motif has a PO backbone (43, 68). The addition of poly G motifs to the 50 and 30 ends of the ODN enhance the cellular uptake and substantially improve its ability to activate NK cells (43, 83) and to induce IFN-α production from human plasmacytoid DC precursors (81). These elements define the class of CpG-A ODN: poly G motifs with PS linkages at the 50 and 30 ends and a PO palindromic CpG-containing sequence in the ODN center (81). CpG-A ODN are generally similar to bDNA in their strong activation of NK cells, but the IFN-α expression induced by an optimal CpG-A ODN is far higher, and the level of B cell activation generally lower, than that induced by the bDNA. Even in comparing bDNA from different bacteria, there are striking differences in the levels of immune activation, with E. coli DNA tending to be much more stimulatory than that from C. perfringens or Streptococcal or Staphylococcal sp (10, 11, 32, 84). The poly G motifs of CpG-A ODN give them excellent cell uptake, but the uptake of bDNA or PO ODN without poly G motifs can be similarly enhanced using cationic lipids (85) or antibodies (80), in which case the cytokine induction is increased to a level comparable to that seen with CpG-A.

Characteristics of CpG-B ODN In contrast to CpG-A ODN, CpG motifs in nuclease-resistant PS backbones have dramatically enhanced B cell stimulatory properties but reduced NK stimulation, despite being much more stable than PO ODN (43, 70, 73, 86). CpG-B ODN as

1 Feb 2002

9:54

AR

AR152-22.tex

AR152-22.SGM

LaTeX2e(2001/05/10)

P1: GJC

CPG MOTIFS

717

Annu. Rev. Immunol. 2002.20:709-760. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

a class have a fully PS-modified backbone with one or more CpG motifs and no poly G motif (82). Even though PS ODN as a family have similar properties, there are still qualitative differences in their effects, and some are substantially more effective at inducing the expression of TNF-α than others (87). To some degree then, every DNA molecule containing CpG motifs must be considered as a separate agent.

Identification of Methylated CpG Motifs and CpG-N Motifs That Neutralize the Effects of Immune Stimulatory CpG Motifs Despite CpG suppression and methylation, vertebrate DNA contains many unmethylated CpG motifs, which suggests that high enough concentrations of vertebrate DNA may be immune stimulatory. However, unless vertebrate DNA is transfected into cells (17), it is nonstimulatory (4, 11, 14). Indeed, addition of equal amounts of vertebrate DNA to bDNA abolishes the CpG-induced IFN-γ secretion and B cell activation, which demonstrates that the vertebrate DNA must contain inhibitory sequences or structures (88). Using plasmids in which the number and type of CpG motifs were manipulated, we demonstrated that methylation of CpG motifs converts a plasmid into an inhibitor of immune activation by unmethylated CpG motifs (88). Vertebrate DNA that has been almost completely unmethylated still has no immune stimulatory effects when added into cultures of CpG-responsive cells, which demonstrates that it must contain immune neutralizing sequences that block the effects of the unmethylated CpG motifs present (32). Genomic sequence analysis showed that the human genome strongly suppressed stimulatory CpG motifs (such as GACGTT) but not CpG motifs in which the CG was preceded by a C and/or followed by a G (29). These neutralizing motifs (CpG-N motifs) specifically antagonize the effects of the stimulatory CpG motifs (28). Poly G sequences in a PS ODN are also extremely effective CpG-N motifs, and they block the ability of immune stimulatory CpG motifs to activate NF-κB (89). DNA vaccines typically contain several hundred CpG motifs, some of which are in an immunostimulatory context, and others of which are CpG-N motifs. Some investigators have reported that addition of just two immune stimulatory CpG motifs to a DNA vaccine can enhance its ability to induce an antigen-specific immune response (90), but most investigators have found that minor changes in the number of CpG motifs in a plasmid do not substantially alter its function (91, 92). To determine whether the efficacy of a DNA vaccine may be improved by deleting some of the CpG-N motifs, we performed in vitro mutagenesis on a DNA vaccine, eliminating 52 of the 134 CpG-N motifs that were present, and we compared this to the parent DNA vaccine as well as to another vaccine in which 16 or more stimulatory CpG motifs were added after deletion of the 52 CpG-N motifs. These vectors showed progressively increasing immune stimulatory effects with the deletion of the CpG-N motifs and further addition of 16 CpG motifs; however,

1 Feb 2002

9:54

AR

Annu. Rev. Immunol. 2002.20:709-760. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

718

AR152-22.tex

AR152-22.SGM

LaTeX2e(2001/05/10)

P1: GJC

KRIEG

addition of further stimulatory CpG motifs actually resulted in a fall of the antibody response to the DNA vaccine (28). Perhaps the addition of too many CpG motifs to a DNA vaccine may actually lead to the production of high levels of type I interferons with resultant inhibition of plasmid expression and loss of immunogenicity. In the gene therapy field, delivery of plasmid vectors has been associated with highly undesirable CpG-induced immune stimulatory effects that can be reduced but not eliminated by the use of steroids or chloroquine, which blocks the CpG signal transduction pathways (see below) (93, 94). In preliminary studies, the immune stimulatory effects of gene therapy vectors can be significantly reduced by deletion of CpG motifs from the vector, but we have not been able to show any further reduction in immune stimulatory effects by addition of CpG-N motifs to the vectors. These data suggest complex interactions between CpG-S and other DNA sequences in the plasmids and suggest the need for further studies into the regulation of immune stimulation by CpG DNA and other sequences.

CELLULAR IMMUNOLOGY OF CpG DNA AND INDUCTION OF TH1-TYPE IMMUNE RESPONSES B Cells Optimal CpG-B ODN are extraordinarily strong mitogens for B cells from essentially all vertebrates, including the mouse, human, cow, sheep, cat, dog, goat, horse, pig, chicken, rat, and rabbit (4, 30, 38, 39, 42, 48). Both bDNA and CpG-B ODN induce B cells to enter the G1 phase of the cell cycle and secrete IL-6 and IL-10 within a few hours (95, 96), but the poly G motifs of CpG-A ODN appear to have an inhibitory effect, reducing the level of B cell activation. The CpGinduced IL-6 expression is required for the B cells to proceed to secrete IgM (95). CpG-induced IL-10 production functions as a counter-regulatory mechanism to reduce the magnitude and duration of IL-12 secretion (97). Cyclosporin blocks the CpG-induced B cell secretion of IL-10 but does not block the CpG-induced macrophage production of IL-12 (96). By preventing the “off” signal from IL-10, Cyclosporin causes a doubling in the IL-12 response to CpG in vitro and in vivo (96). In vivo, CpG DNA induces NK cells to secrete IFN-γ (98), which enhances the B cell response, as mice genetically deficient in IFN-γ produce less than half of the usual IL-6 and IgM response to CpG DNA (99). In addition to secreting cytokines and Ig, CpG-activated B cells express increased levels of the Fcγ receptor and costimulatory molecules such as class II MHC, CD80, and CD86 (4, 100, 101). Human B cells upregulate expression of not only these costimulatory molecules but also CD40 and CD54 (39). CpG DNA also activates malignant B cells from patients with chronic lymphocytic leukemia, driving them into the cell cycle, inducing cytokine secretion, and upregulating surface expression of CD40, CD58, CD80, CD86, CD54, and MHC class I (102). In mature peripheral B cells, low concentrations of CpG DNA strongly synergize with signals through the BCR, leading to an approximate tenfold increase in B cell proliferation and antigen-specific Ig secretion and IL-6 secretion (4). The

1 Feb 2002

9:54

AR

AR152-22.tex

AR152-22.SGM

LaTeX2e(2001/05/10)

P1: GJC

Annu. Rev. Immunol. 2002.20:709-760. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

CPG MOTIFS

719

synergy is so strong that multivalent cross-linking of the BCR can synergize with even methylated CpG ODN and non-CpG ODN to induce B cell proliferation and Ig secretion, though with less efficacy compared to CpG ODN (103). The synergy between CpG and the BCR is evident as early as the induction of MKK3, MKK4, MKK6, and JNK, but there is no synergy for activation of ERK (104) (A. K. Yi and A. M. Krieg, manuscript in preparation). BCR ligation in certain B cell lines such as WEHI-231 or BKS-2 triggers BCR-induced apoptosis, which is opposed by CpG DNA (105–107). This antiapoptotic effect of CpG DNA is associated with increased NFκB p50/c-Rel activity and maintenance of c-myc levels (105, 106). CpG-B ODN or even ODN with a PO backbone also have antiapoptotic activity on isolated mature primary B cells, which prevents their normal spontaneous apoptosis in tissue culture by maintaining NFκB expression (37) and by blocking the spontaneous fall in the mitochondrial membrane potential (108). Like the other stimulatory activities of CpG DNA on B cells, these protective effects are completely reversed by poly G motifs, such as those in CpG-A ODN (89). CpG DNA has also been reported to protect B cells against Fas-mediated apoptosis by downregulating Fas expression on B cells stimulated through CD40 (109).

Dendritic Cells Distinct subsets of murine and human dendritic cells (DC) and DC precursors express different subsets of TLRs that enable them to induce different patterns of immune responses to different pathogens. Among human DC precursors, the plasmacytoid pre-DC (pDC, sometimes called DC2) strongly express TLR9, while myeloid CD11c+ pre-DC do not, instead expressing different TLRs such as TLR4 that cause them to be activated by a different set of pathogen molecules including LPS (110, 111). Among human DC subsets, so far only the pDC are clearly demonstrated to be directly activated by CpG DNA, which induces them to have growth factor–independent survival in culture, resistance to IL-4–induced apoptosis, increased surface expression of MHC class II, ICAM-1 and the costimulatory molecules CD40, CD54, CD80, and CD86, cytokine secretion (IL-6, TNF-α), IFN-α secretion, chemokine production (IL-8, IP-10, GM-CSF), and maturation to become CD83 bright with increased activation of allogeneic T cells (49, 78, 110, 111). In fact, CpG-B ODN alone could substitute for the cytokine requirement for survival of these primary DCs, and they were even superior to GMCSF in promoting pDC survival in tissue culture (49). The two factors acted synergistically together, which shows that they work through distinct pathways. In combination, CD40 ligation and CpG DNA induce synergistic IL-12 p70 production in human plasmacytoid pre-DC (110). pDC are also notable for their role as the primary source of the IFN-α that is rapidly produced in response to viral infections. It is surprising that the two classes of CpG ODN have quite different activities upon plasmacytoid pre-DC IFN-α production: optimal CpG-A ODN induce extraordinarily high levels of IFN-α production, but CpG-B ODN induce relatively little, despite similar abilities to induce DC maturation and TNF-α production (81).

1 Feb 2002

9:54

Annu. Rev. Immunol. 2002.20:709-760. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

720

AR

AR152-22.tex

AR152-22.SGM

LaTeX2e(2001/05/10)

P1: GJC

KRIEG

Perhaps the most commonly studied type of DC is the monocyte-derived DC, which comes from culturing human peripheral blood monocytes for approximately five days in GMCSF + IL-4. These DCs, which do not express TLR9, do not express costimulatory molecules in response to CpG-B ODN (49) but have been reported to express IL-12 and IL-18 mRNA (112), though this result has not been repeated. Human monocyte-derived DC have also been reported to be activated by CpG-A DNA in the form of plasmids, but this result has to be considered cautiously because of the marked LPS sensitivity of these cells (113, 114). Similar to human pDC, murine bone marrow–derived DCs and Langerhans cells are activated by CpG-B ODN to express costimulatory molecules, but they also downregulate expression of E-cadherin and α 6 adhesion molecules and secrete IL-12 and IL-6 (115, 116, 117). These changes in adhesion molecule expression are coupled to the appearance of filopods and loss of adherence. It is interesting that these direct effects of CpG DNA differ somewhat from LPS in that, while LPS induces murine DCs to make large amounts of TNF-α and relatively low amounts of IL-12, CpG DNA activates the opposite pattern of cytokine secretion with lower levels of TNF but higher levels of IL-12 (115). LPS upregulates DC expression of CD40 to a degree similar to CpG DNA, but only CpG efficiently primes the DC for high-level IL-12 p70 production in response to CD40 crosslinking (118). In combination, CD40 ligation and CpG DNA induce synergistic IL-12 p70 production in murine splenic DC, including both the CD8α + and CD8α − subsets (118). These stimulatory effects of CpG DNA enhance the ability of the DC to activate allogeneic T cells (115, 116). Subcutaneous administration of CpG DNA also leads to in vivo activation of skin Langerhans cells to upregulate costimulatory molecules and produce IL-12 (115). These CpG-activated Langerhans cells change morphology and migrate out of the skin within 2 h (117). CpG injection also causes changes in the localization of splenic DC that disappear from the marginal zone and T cell areas by 48 h after injection (119). In summary, CpG DNA induces a pattern of Th1-like immune activation in human and murine DC and DC precursors and activates their migration.

Monocytes and Macrophages Purified human monocytes do not express TLR-9 and are not activated by CpG ODN (111). Nevertheless, CpG DNA treatment of human PBMC or whole blood secondarily activates the monocytes to express increased levels of CD40 and CD69 and to produce IL-6 and TNF-α, though with delayed kinetics compared to LPS (120, 121). Even without CpG motifs, the PS ODN backbone can enhance LPSstimulated monocyte TNF production (63, 122). Among monocyte-related cells, microglial cells and astrocytes are also activated by CpG-B ODN (123, 124). In murine macrophage/monocytes CpG DNA causes the direct activation of NFκB (86) and initiation of cytokine expression including TNF-α (116). For example, macrophage expression of the inducible nitric oxide synthase and production of nitric oxide in response to CpG DNA is not direct but requires IFN-γ priming (125). CpG DNA can synergize with IFN-γ in driving expression from an HIV-1 long

1 Feb 2002

9:54

AR

AR152-22.tex

AR152-22.SGM

LaTeX2e(2001/05/10)

P1: GJC

Annu. Rev. Immunol. 2002.20:709-760. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

CPG MOTIFS

721

terminal repeat reporter construct (125). The macrophage-like cell line RAW 264 has been used to better understand the mechanism of the IL-12 secretion induced by CpG DNA and the role of IFN-γ in enhancing this. Both CpG DNA and LPS activate an IL-12 p40 promoter-reporter construct, and this response is increased two- to fivefold by IFN-γ (126). In this cell line, both LPS and CpG DNA induce similar NFκB activation, although only CpG is a strong activator of IL-12 p40 transcription. At low concentrations, CpG DNA and LPS show synergy for inducing macrophage nitric oxide production and monocyte cytokine production (120, 127). This synergy is reportedly regulated in part at the level of NFκB (128) and in part at a posttranscriptional level (129). CpG-A and -B ODN induce RAW 264 and spleen cells to express increased levels of cyclooxygenase-2 and to secrete PGE2 (130). This PGE2 has a counter-regulatory effect on the CpG-induced Th1 state by inhibiting IFN-γ production. Mice injected with a CpG-B ODN together with a cyclooxygenase-2 inhibitor do not suffer from the inhibitory effects of the PGE2 and have a substantial increase in the CpG-induced serum IFN-γ response (130). After the intial positive effects of CpG DNA on antigen processing and presentation, it downregulates macrophage class II MHC antigen processing and presentation within about 18 h (131, 132). ODN can have backbone-dependent effects on macrophages—PS but not PO ODN have chemoattractant effects on primary macrophages (133). The chemoattractant effect of PS ODN does not require a CpG motif but is enhanced by poly G motifs (133). CpG DNA also induces human T cells to produce a heat labile factor that inhibits macrophage differentiation and adherence (134).

Natural Killer Cells Since the initial observation that mycobacterial DNA activates murine NK cells to make IFN-γ and to have increased lytic activity (135), many additional studies have confirmed and extended this finding for mouse (9, 11, 12, 136–138) and human cells (139). As reviewed above, these investigators concluded that palindromic sequences in mycobacterial DNA were responsible for the NK activation. The presence of a palindrome in an ODN probably leads to the formation of a double-stranded duplex region that stabilizes the structure against degradation compared to an ODN without a duplex. Palindromes are not required for NK stimulation but simply CpG motifs in appropriate base contexts are required, and the stimulatory effects are dramatically enhanced if the ODN is stabilized against nuclease degradation by modifying the ends with PS linkages (43). However, if the CpG motifs themselves are modified with PS linkages, rather than just the ends, then the magnitude of activation is much less than that with the same sequence in a PO backbone (43, 70). The major source of the initial IFN-γ production in CpG-stimulated murine spleen cells appears to be NK cells (98). The stimulatory effects of CpG DNA on murine NK cells are not direct but require either the presence of adherent cells or their CpG-conditioned supernatants, which contain IL-12, TNF-α, and type I interferons (43, 98). Thus, CpG DNA appears to act as a costimulatory signal for murine NK cells. ODN containing

1 Feb 2002

9:54

AR

722

AR152-22.tex

AR152-22.SGM

LaTeX2e(2001/05/10)

P1: GJC

KRIEG

CpG palindromes have been reported to directly activate human NK cells (40), but it is uncertain whether small numbers of contaminating pDC may have been present in the cell preparation and contributed to this activation.

Annu. Rev. Immunol. 2002.20:709-760. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

T Cells Like NK cells, CpG DNA generally has not been reported to have direct stimulatory effects on resting T cells (40, 74, 75, 140). In murine mixed cell populations, the type I interferons produced by CpG-stimulated adherent cells stimulate T cells to produce some activation and costimulatory molecules but also inhibit their proliferative response to TCR ligation (74). However, if T cells are highly purified away from adherent cells, then CpG DNA synergistically enhances the proliferative response of murine T cells to TCR ligation (141). These costimulatory effects were also observed in T cells from mice lacking CD28, raising the possibility that CpG DNA may substitute for normal T cell costimulatory signals (141). Recent studies by the same investigators have led to a reinterpretation of these experimental results. The murine cell costimulatory effect is now reported to be due not to CpG motifs but to a second type of DNA motif in these PS backbone ODN, poly-G (G-quartets), consisting of 4 Gs in a row, or 2 or more regions with 3 Gs in a row (57, 142) The question of the possible costimulatory effects of various ODN on T cells has recently become even more complicated. Iho et al. have recently reported that PO ODN containing certain CpG motifs (CpG-A type) directly stimulate purified human T cells to proliferate, as long as the T cells are activated through the TCR with a strongly cross-linking stimulus such as magnetic beads coated with antiCD3 (40). This response did not require the presence of IL-2. It is surprising that some G-rich sequences that were not strong stimulators of human NK cells, such as the motifs CGGCCG, GCGCGC, and TCGCGA, were quite effective at costimulating T cells to secrete IFN-γ (40). This suggests that NK and T cells may detect different subsets of CpG motifs and may express different CpG receptors. Stimulation of these human T cells was abolished if the cytosine in the CpG motif was replaced by 5-methyl cytosine. PO ODN containing poly-G sequences did not show any T cell costimulatory properties in these studies with human T cells, which raises the question of whether the costimulatory effect observed by Wagner and colleagues is limited to murine T cells or to ODN with a PS backbone. On the other hand, Kranzer et al. have reported that PS CpG ODN failed to costimulate purified human T cells activated with anti-CD3 or anti-TCR Ab (143). In whole PBMC, the CpG had strong stimulatory effects that were dependent on the production of type I IFN and IL-12. The indirect effects of CpG on T cell activation are discussed in more detail below in the section on CpG as a cancer vaccine adjuvant.

Neutrophils CpG does not induce purified neutrophils to express activation markers or to have an enhanced oxidative burst (144). However, CpG-B ODN indirectly activate murine

1 Feb 2002

9:54

AR

AR152-22.tex

AR152-22.SGM

LaTeX2e(2001/05/10)

P1: GJC

CPG MOTIFS

723

Annu. Rev. Immunol. 2002.20:709-760. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

neutrophils. Administration of a CpG-B ODN to mice causes a transient decrease in peripheral blood neutrophils, presumably secondary to changes in adhesion molecule expression and redistribution into tissues (G. Lipford, personal communication). Weighardt et al. administered a CpG-B ODN to mice and then challenged them with a peritoneal infection. CpG treatment led to an enhanced neutrophil influx into the site of infection, and these neutrophils were phenotypically activated, had enhanced phagocytosis, and increased production of reactive oxygen species (ROS) (145). Thus, although CpG DNA does not appear to activate neutrophils directly, it improves their ability to provide effective host defense.

Nonimmune Cells The vast majority of published studies on CpG DNA have explored its effects on various immune cell populations. However, a few investigators have reported activation of nonimmune cells. Carlow et al. have described CpG-induced stimulation of L cells, which are of stromal origin, to produce IFN-α/β upon transfection with plasmid DNA (146). This response was not seen with LPS and required the use of lipids to transfect the DNA into the cells, leaving open the question of whether it would occur under physiologic conditions. Moreover, the investigators did not include a negative control of methylated or vertebrate DNA. Thus, it remains possible that the observed effect may be CpG-independent, perhaps owing simply to transfection with double-stranded DNA, which induces several signaling pathways (17). Bacterial DNA or a CpG ODN have also been reported to induce human gingival fibroblasts to activate NFκB and secrete IL-6 (147). In this case, the effect did not occur with eukaryotic or methylated DNA, which suggests that it is CpG-specific. Until further studies are reported, a certain degree of caution may be applied to the interpretation of these results because the IL-6 response is notoriously sensitive to LPS, which was not rigorously excluded as a contributory factor to the results. At the time of writing, the only cells that CpG DNA directly activate upon exposure to CpG DNA are the TLR-9 expressing cells, B cells and pDC (110, 148) (see below under Mechanisms of Action). TLR-9 appears to be both necessary and sufficient for cells to respond to CpG DNA, and there may be physiologic or pathologic conditions where TLR-9 would be expressed in nonimmune cells, in which case they would be expected to become CpG responsive.

Creation of a Th1-Like Cytokine Milieu As summarized above, the predominant effects of CpG DNA in vitro and in vivo are Th1-like. Systemic injection of CpG DNA creates a systemic Th1-like response (149, 150). Likewise, local injection of CpG subcutaneously into the footpad of a mouse induces IL-12 and IFN-γ production in the draining lymph nodes, with lymphadenopathy that peaks at 7–10 days (142). DC become a more prominent population in the lymph nodes and exhibit an activated phenotype with increased expression of costimulatory molecules (151). This local Th1-like environment is sustained enough so that mice can resist a local leishmania challenge

1 Feb 2002

9:54

724

AR

AR152-22.tex

AR152-22.SGM

LaTeX2e(2001/05/10)

P1: GJC

KRIEG

for two weeks after a single dose and respond to an antigen injection with a Th1-biased response including CTL even five weeks later (142). Intradermal or intranasal delivery of the CpG also results in a localized state of Th1 predisposition (101).

MECHANISMS OF ACTION

Annu. Rev. Immunol. 2002.20:709-760. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

Cellular Binding and Uptake Cell surface DNA binding proteins are well described but appear to lack any sequence specificity; stimulatory CpG ODN and nonstimulatory ODN bind equally well to cell membranes (4, 44, 152). The only exception to this nonsequencespecific binding is that ODN containing G quartets show enhanced binding to lipid bilayers and cells (83, 153, 154). In theory, a surface protein could bind DNA in a sequence-independent fashion but transduce a signal in a CpG-dependent fashion. Our initial studies in murine cells showed that CpG ODN immobilized on a solid support could not activate lymphocytes, which suggests that cell uptake was required (4). In contrast, Lipsky and colleagues have reported that human B cells are stimulated by CpG ODN immobilized on sepharose beads (48, 155). Although these experiments suggest that CpG ODN may work through a cell surface receptor, sepharose beads can still be taken up by cells in tissue culture, leaving open the possibility that CpG ODN may still work through an intracellular signaling pathway rather than a cell surface receptor (156). CpG ODN that were linked to latex, magnetic, or gold beads could not be taken up and lost their stimulatory activity (156). Furthermore, lipofection of ODN into spleen cells enhances the immune stimulatory effects even for hexamer ODN, which normally are nonstimulatory (44, 85, 157, 158). Curiously, CpG-induced IFN-γ secretion is enhanced to a far greater extent than is IL-12 (157). Taken together, these data suggest that although their cellular uptake may involve binding to cell surface proteins, the immune stimulatory effects of CpG ODN most likely require binding to an intracellular (or endosomal) receptor. Different lymphocyte subsets have quite different levels of ODN uptake. In vitro, B cells and monocytic cells have the highest rates of ODN uptake, while T cells and neutrophils have low rates with no differences between CD4+ and CD8+ T cells (159–162). Uptake in B cells is regulated depending on the stage of cell differentiation, with the highest uptake observed in the pro– and pre–B cells (53). ODN uptake in both B and T cells is highly inducible by mitogens (159, 160). ODN uptake in malignant cells and cell lines is typically higher than that in primary cells and related to cellular activation (160). Cellular uptake of ODN may be regulated by the local concentration of lactoferrin, a protein present at high concentrations in mucosal surfaces, and which inhibits the immune stimulatory effects of CpG-B ODN (163). Dead or apoptotic cells have extremely high levels of ODN, which are located in the nucleus and can confound the interpretation of ODN uptake experiments (45).

1 Feb 2002

9:54

AR

AR152-22.tex

AR152-22.SGM

LaTeX2e(2001/05/10)

P1: GJC

Annu. Rev. Immunol. 2002.20:709-760. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

CPG MOTIFS

725

The mechanism(s) of ODN uptake in lymphocytes remain uncertain. ODN are large polyanions and cannot diffuse across cell membranes. Only a single protein has been demonstrated to mediate DNA uptake: a 45-kDa protein expressed in kidney brush border membranes acts as a voltage-gated channel for the entry of ODN into these cells (164). There is no evidence that this protein is expressed in lymphocytes. ODN uptake in lymphocytes is an active process that is temperature- and energy-dependent, competable, saturable, and is generally sequence-independent with the exception that G-rich sequences can enhance the uptake of PO ODN (83, 165). Pinocytosis facilitates ODN uptake at relatively high concentrations (<1 µM), whereas receptor-mediated endocytosis, possibly clathrin-dependent, appears to be most important at lower ODN concentrations (166, 167). In living and unfixed cells, ODN are localized in the endosomes (25, 45, 165, 168). The results of localization studies using fixed cells depend critically on the fixation protocol used, with mild paraformaldehyde fixation appearing to give the most accurate results (168). If ODN are injected directly into the cytoplasm, they rapidly localize in the cell nucleus (169–171). Indeed, measures to enhance the endosomal release of ODN such as delivery with cationic liposomes, conjugation to cholesterol, or the use of endosome-disrupting agents, enhance not only the antisense efficacy, but also the immune stimulatory effect of CpG ODN (25, 44, 172, 173). The in vivo uptake and pharmacokinetics of ODN have been studied in detail, mostly by investigators developing antisense ODN (reviewed in 174). PS ODN delivered by IV or SC administration mainly accumulate in the liver and kidney, with lower levels in the spleen and bone marrow (174). PS ODN are quite stable in vivo with a tissue half-life of about 48 h, but PO DNA is rapidly degraded with a half-life of approximately 5 min (175). The cellular distribution of PS ODN after IV administration has been examined using FITC-conjugated ODN, demonstrating heterogeneous uptake among PBMC, spleen, and bone marrow cells, with highest levels of uptake in monocytes and macrophages, intermediate levels in B cells, and low levels in T cells (176).

Endosomal Acidification/Maturation Following their uptake into lymphocytes, ODN appear to be located within the endosomal compartment (53, 174). Much evidence points to the endosome as a potential site of CpG-induced signal initiation. We showed that monensin, chloroquine, and bafilomycin A, which interfere with endosomal acidification and/or maturation, completely block the immune stimulatory activity of CpG DNA at low concentrations that did not inhibit LPS, anti-CD40, cross-linking of the B cell antigen receptor, or phorbol 12-myristate 13-acetate (177). Other compounds structurally related to chloroquine, such as quinacrine, are even more potent antagonists of CpG-induced immune stimulation (178, 179). These compounds must act at an extremely early step in the CpG-induced signaling pathway because their inhibitory effects are already apparent by 5 min and because they block all of the signaling pathways yet known to be induced by CpG (104, 148, 152, 177).

1 Feb 2002

9:54

726

AR

AR152-22.tex

AR152-22.SGM

LaTeX2e(2001/05/10)

P1: GJC

KRIEG

Annu. Rev. Immunol. 2002.20:709-760. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

TLR-9, A Link Between CpG DNA and the Activation of Cell Signaling Pathways Toll-like receptors (TLRs) function as pattern recognition receptors (PRRs) to initiate innate immune activation in response to infection by detecting pathogenspecific molecular structures (2, 5). Immune activation by CpG-B ODN depends on TLR-9, as mice genetically deficient in this molecule show no CpG-induced activation of B cells, DC, or NK cells (180). The role of TLR-9 in mediating the effects of other forms of CpG DNA, such as CpG-A ODN or bDNA, is not yet clear. TLR-9 has a transmembrane domain and other features of a type I membrane protein (181, 182). Expression of TLR-9 in human embryonic kidney cells, which normally are not activated by a CpG-B ODN, causes them to become CpGresponsive (148). Therefore, expression of TLR-9 is both necessary and sufficient for CpG-induced NFκB. Among human cell types, TLR-9 expression is highest in B cells and plasmacytoid DC, which are also the only cell types that have been reproducibly shown to be directly activated by PS CpG ODN (110, 148). Perhaps the most direct evidence for the role of TLR-9 in the direct recognition of a CpG motif is the fact that 293 cells transfected to express the mouse TLR-9 protein become optimally responsive to the preferred mouse CpG motif, GACGTT, while 293 cells transfected to express the human TLR-9 protein become optimally responsive to the preferred human CpG motif, GTCGTT (148). Thus, the TLR-9 protein determines the species specificity of CpG motifs. These data are consistent with the view that TLR-9 is an essential component of the postulated CpG DNA receptor and that it acts upstream of the adapter protein MyD88 and links CpG motif recognition to the TLR/IL-1R signaling pathway. Circumstantial evidence suggests that TLR-9 may be present in the endosomes, where it could interact with CpG DNA (180). Recent studies indicate that TLR-9 expression enhances the uptake of CpG DNA but not non-CpG DNA (183), which would seem to be inconsistent with previous studies demonstrating no difference in the rate of uptake of CpG vs. non-CpG DNA (see above). It is likely that ODN are taken up by cells through multiple pathways, and it remains possible that TLR-9 may be directly or indirectly involved in these processes.

Activation of the Mitogen-Activated Protein Kinase Pathways Mitogen-activated protein kinases (MAPKs), such as extracellular receptor kinase (ERK), p38, and the c-Jun NH2-terminal kinase (JNK), mediate leukocyte responses to diverse stimuli. B cells activated by CpG DNA show activation of both the p38 and JNK pathways within seven minutes (104). In comparison to B cell activation by CD40 ligation, CpG DNA induced a slightly slower onset of MAPK phosphorylation but a longer duration of activation. It is interesting that the JNK isoforms phosphorylated in response to CpG DNA apparently had identical molecular weights to those phosphorylated in response to CD40 ligation but were different from those seen in B cells activated through the BCR (104). Human B

1 Feb 2002

9:54

AR

AR152-22.tex

AR152-22.SGM

LaTeX2e(2001/05/10)

P1: GJC

Annu. Rev. Immunol. 2002.20:709-760. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

CPG MOTIFS

727

cells activated by CpG DNA also show activation of the p38 and JNK MAPKs but not ERK (39). The p38 pathway appears to be required for CpG-induced B cell cytokine secretion, since this secretion is completely blocked by pretreatment of the cells with a p38 inhibitor (104). The p38 and JNK pathways are also rapidly activated in macrophages following exposure to CpG DNA (104, 152). Of note, CpG also activates the ERK pathway in primary macrophages and RAW264.7, a macrophage-like cell line, and this contributes to CpG-induced TNF production but has a negative feedback effect on the IL-12 p40 promoter, which results in decreased release of IL-12 (184). These studies suggest that the activation of p38 and JNK MAPKs by CpG DNA contributes to IL-12 production, but ERK activation can exert a negative feedback effect. In a different macrophage-like cell line, J774, and in murine DCs, no CpGinduced ERK activity was detectable, and therefore there is no negative feedback effect of this type. The MAPK activation following CpG-B treatment of J774 cells has been reported to contribute to the induction of another counter-regulatory pathway mediated by suppressors of cytokine signaling (SOCS) (185). SOCS-1 and -3 mRNA expression is induced by CpG-B ODN and functions to inhibit IFN-γ signaling (185).

Activation of Nuclear Factor κB Even prior to the discovery of the CpG motif, McIntyre et al. reported that in an intended antisense experiment, a control sense ODN to the p65 subunit of NFκB caused an unexpected upregulation of NFκB activity in murine B cells (22). Further studies revealed a CpG motif in this immune stimulatory ODN that was responsible for the activation of NFκB (A. M. Krieg & R. Narayanan, unpublished data). Both CpG-A and -B ODN trigger the degradation of IκBα and IκBβ and the activation of NFκB in both macrophages and murine and human B cells (39, 86, 106, 186). This requires IKKβ, which is reportedly induced secondarily to CpG-induced activation of DNA-dependent protein kinase (DNA-PK) (187). CpG DNA induces the production of reactive oxygen species (ROS) within 5 min, which also contributes to NFκB activation (95, 106, 177). In B cells, the dominant form of NFκB induced by CpG DNA appears to be a p50/c-Rel heterodimer, while in macrophages it appears to be a p50/p65 heterodimer (106, 186). This NFκB activation results in enhanced transcriptional activity from the human immunodeficiency virus long-terminal repeat (86) and is required for the CpG-induced protection of B cells against apoptosis (106, 108).

Activation of Transcription and Translation The MAPK pathways activated by CpG DNA have been reported in other systems to lead to the activation of multiple transcription factors, including ATF-1, ATF-2, the cyclic AMP response element binding protein (CREB), Elk-1, Max, and c-Jun (188, 189, 190). Some of these factors become phosphorylated and activated in response to CpG DNA treatment of B cells and/or macrophages (104, 152). Like

1 Feb 2002

9:54

Annu. Rev. Immunol. 2002.20:709-760. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

728

AR

AR152-22.tex

AR152-22.SGM

LaTeX2e(2001/05/10)

P1: GJC

KRIEG

NFκB, which is also activated by CpG DNA and B cell types, these transcription factors are important regulators for the expression of many cellular protooncogenes and proinflammatory cytokines. CpG DNA also induces increased mRNA levels of several other transcription factors, including c-myc, Ets-2, C/EBP-β, and C/EBP-δ (177, 191). Promoter activity for IL-6 and HIV is induced by CpG DNA in murine macrophages and B cells (95, 191) and in a human myeloma cell line (192). This transcriptional induction appears to be extremely rapid, with increased levels of mRNA within 15 minutes (37). Among the genes whose RNA expression is increased in B cells are myc, myn, EGR-1, Jun, Bcl-2, Bcl-xL, IL-6, IL-10, and IL-12 (37, 95, 96, 105, 106). The CpG-induced activation of the IL-12 p40 promoter has been attributed to induction of the NFκB p50/c-rel heterodimer and to C/EBPβ (193). It is interesting that although the overall effect of CpG DNA is the strong promotion of Th1-like immune responses, the B cell production of IL-10 acts to reduce the level of IL-12 secretion induced by CpG DNA (96). CpG DNA also has potent transcription-activating effects on macrophages and DC, which leads to the increased transcription of TNF-α, IL-1β, plasminogen activator inhibitor-2, IL-6, IL-12, Type I interferons, and several costimulatory and antigen presenting molecules such as class II MHC, CD80, CD86, CCR7, and CD40 (74, 86, 97, 110, 115, 186, 194). As reviewed above, NK cells are induced by CpG DNA to produce IFN-γ . Other cytokines whose expression is induced by CpG DNA, but for whom the cellular sources have not yet been conclusively determined, include IL-1RA, MIP-1β, MCP-1, IP-10, IL-15, and IL-18 (110, 195–197).

THERAPEUTIC APPLICATIONS OF CpG DNA Activation of Innate Immune Defenses Against Infection Killed bacteria or their products trigger innate immune responses that increase nonspecific resistance to some infections (198, 199, 200). Presumably, these defensive responses are triggered by host pattern recognition receptors (PRRs), which recognize foreign molecular structures and trigger defensive responses. To investigate the possibility that CpG may be such a molecular structure, we used a model in which BALB/c mice are challenged with approximately 10 LD50 of Listeria monocytogenes, to which these mice are highly susceptible. BALB/c mice pretreated with as little as 0.3 µg of either bacterial DNA or a CpG-B ODN were protected against the infectious challenge (150). However, the mice had to be given the CpG at least 48 h prior to the lethal challenge to be fully protected: mice that were given the CpG DNA at the same time as the challenge or only 24 h before the challenge were not protected. The CpG-induced protection continued for two to four weeks, was associated with increased serum IL-12, and required the presence of IFN-γ (150). These results have been extended to another intracellular bacteria, Francisella tularensis (201, 202). Protection led to the development of memory responses because animals surviving lethal challenge were protected against later

1 Feb 2002

9:54

AR

AR152-22.tex

AR152-22.SGM

LaTeX2e(2001/05/10)

P1: GJC

Annu. Rev. Immunol. 2002.20:709-760. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

CPG MOTIFS

729

rechallenge. Remarkably, protection can be maintained for at least four months with no evidence of desensitization by repeated dosing of mice with CpG at twoweek intervals (203, 204). By encapsulation of the CpG-B ODN within cationic liposomes, the duration of complete protection against challenge with 1000 LD50 of L. monocytogenes could be extended from two to at least four weeks after a single dose (158). Mice treated with CpG DNA are also protected against challenge with Leishmania major and malaria through IFN-γ , nitric oxide, and IL-12–dependent mechanisms (87, 205–208). In contrast to the requirement for pretreatment with CpG DNA in the L. monocytogenes model, mice could be treated with CpG DNA on the day of challenge without loss of protection against either pathogen, and CpG-B ODN could still cure L. major–infected BALB/c mice even when administered as late as 15 days after infection. CpG DNA treatment also protects mice against challenge with Ebola virus or anthrax (209). The protective effect of CpG-induced innate immune activation also extends to polymicrobial sepsis. Weighardt et al. induced intra-abdominal sepsis in mice through colonic puncture and stenting, which causes death within 4–7 days in approximately 50% of control mice (145). However, administration of a single dose of CpG ODN given 6 days before the procedure reduced mortality to about 15%. Moreover, the CpG-treated mice had a dramatically enhanced protective response to the infection, as demonstrated by a more than threefold increase in the number of abdominal neutrophils by 20 h postinfection and by a marked increase in the expression of activation markers and the oxidative burst activity of these cells (145). In addition to providing nonspecific protection against infectious challenge, CpG ODN have a profound effect on hematopoietic function. Even before the identification of the CpG motif, several investigators using antisense ODN noted the induction of sequence-specific extramedullary hematopoiesis and induction of hematopoietic colony formation (22, 210). More recently, these effects were shown to be CpG specific (211). Mice treated with high doses of immune stimulatory PS CpG ODN develop massive splenomegaly and increased spleen granulocytemacrophage colony forming units (GM-CFUs) and early erythroid progenitors. CpG had a radio-protective effect in accelerating the recovery of GM-CFUs and enhancing resistance to listeria infection in irradiated mice (211).

Role of CpG DNA as a Vaccine Adjuvant and in DNA Vaccines Several recent reviews have covered the potent adjuvant activity of CpG DNA in vaccination with various antigens and with DNA vaccines, which has become well established (212). Because of severe space and length limitations, we will therefore provide only the most cursory description of this topic. Briefly, CpG DNA is a stronger Th1-like adjuvant for inducing B cell and T cell responses than the gold standard, complete Freund’s adjuvant, as measured by its ability to drive the differentiation of CTL and IFN-γ secreting T cells. Moreover, CpG DNA can

1 Feb 2002

9:54

Annu. Rev. Immunol. 2002.20:709-760. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

730

AR

AR152-22.tex

AR152-22.SGM

LaTeX2e(2001/05/10)

P1: GJC

KRIEG

be used for vaccination not only through parenteral routes, but also by means of oral or mucosal vaccination. The ODN activate APC to present antigen more effectively and therefore have to be injected in the same location, but not necessarily at the same time; the antigen can be given a week or more after the CpG injection, and outstanding immune responses to the antigen are still seen (142). CpG-B ODN give potent antibody and CTL responses in humans as well as rodents, as opposed to CpG-A ODN, which give weaker Ab responses but may give even stronger CTL, even in the absence of T cell help. The only class of antigens for which CpG ODN are not highly effective are polysaccharide antigens, though with weak protein epitopes, formulation may be required to achieve the full CpG effect. The author apologizes to the many investigators who have performed outstanding studies in this field for the impossibility of referencing their work in this review.

Cancer Immunotherapeutic Activities of CpG DNA CpG DNA MONOTHERAPY As reviewed above (see History), the antitumor effects of infections have been recognized for hundreds of years, effectively demonstrating that immune activation can result in tumor eradication. This point is consistent with the recently rediscovered role of cancer immunosurveillance, mediated by IFN-γ and lymphocytes, in suppressing tumor growth (213). By inducing IFN-γ production and activating lymphocytes, CpG-induced innate immune activation may enhance cancer immunosurveillance and prevent tumor development. This hypothesis has been tested in mice transgenic for c-MYC expressed from an Ig enhancer; the mice have a 95% rate of lymphoma development within four months of life (214). Chronic treatment of these mice with a CpG-B ODN starting at the time of weaning prevented tumor development in 75% of the mice (214, 215). Thus, despite its effect to activate B cell proliferation and block apoptosis, CpG ODN treatment reduced spontaneous B cell lymphomagenesis in susceptible mice. The efficacy of CpG-A DNA in preventing or treating tumor development or metastasis in mice has been examined in several experimental models. In some models, plasmids containing CpG motifs were stabilized by the formation of complexes with cationic lipids, which enables them to induce systemic NK cell activation and IFN-γ production (216). Methylation of the plasmids reduced their immune stimulatory effects, demonstrating the CpG-specificity. Mice were given tumors by IV injection with experimental fibrosarcoma, melanoma, or colon carcinoma cell lines, and then treated with IV cationic lipid-DNA complexes three and ten days later (216). Mice with intact immune systems had a marked decrease in the number of pulmonary metastases, compared to mice given lipid or DNA alone, and tumor growth in mice with preexisting tumors was slowed (216). However, these CpG-A molecules had no protective effects in mice depleted of NK cells with anti-asialo Gm1 antiserum or in mice genetically deficient in IFN-γ (216). Similar cationic lipid:plasmid DNA complexes are effective in the treatment of mice with established indolent or aggressive mesothelioma in two models, AC29 and AB12 (217). In the less aggressive model, AC29, treatment with CpG-A resulted in a

1 Feb 2002

9:54

AR

AR152-22.tex

AR152-22.SGM

LaTeX2e(2001/05/10)

P1: GJC

Annu. Rev. Immunol. 2002.20:709-760. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

CPG MOTIFS

731

>90% long-term survival rate, while in the more aggressive AB 12 model, treatment still gave >40% long-term survival (217). In both models, CpG treatment led to a protective memory response, because surviving mice resisted repeat tumor challenge. CD8+ T cells and NK cells were absolutely required, and CD4+ T cells contributed to the CpG-induced protective effect (217). Mice injected with either plasmid DNA in liposomes (CpG-A type DNA) or CpG-B ODN are protected against a subsequent IV challenge with fibrosarcoma or thymoma cells through a mechanism that appeared to involve IFN-α secretion by APC, followed by activation of NK and/or NKT cells (218). Systemic or local therapy with a CpG-A ODN protected 80% of syngeneic C57 BL/6 mice from a lethal challenge of B16 melanoma (219). Even when treatment was begun three days after tumor challenge, 60% of mice could still be cured of disease with CpG-A treatment, but CpG-B ODN treatment was less effective. SCID mice were also protected against tumor challenge by CpG DNA, which indicates that neither B nor T cells are required. In contrast to the mesothelioma models described above, CpG-treated mice that survived tumor challenge were not protected against subsequent tumor challenges, which demonstrates that specific immunity was not generated (219). Systemic or local therapy with a CpG-B ODN is more effective than a CpGA ODN in treating a T-cell lymphoma model, which establishes the point that different classes of ODN are optimal in the immune therapy of different tumor types (219). Daily injection of CpG-B ODN for 15 days into syngeneic neuroblastoma tumor nodules results in complete tumor regression in about 50% of the mice (220). Similar results have been reported in a glioblastoma model (221). Animals cured with CpG treatment were protected against further tumor challenge, suggesting the development of an antigen-specific T cell response, in contrast to the lack of a role for T cells in the case of the B16 melanoma model described in the previous paragraph. In the C1498 mouse AML model, a CpG-B ODN had both preventive and therapeutic activity, while the same CpG-A ODN (1585) that was highly effective in the B16 melanoma model described above was ineffective (222). This emphasizes the point that no single method of CpG immunotherapy is optimal in all models. NK cells were required for the CpG-induced protective effect, but neither T nor B cells contributed, and no memory response appeared to be induced. Both CpG-A and CpG-B ODN improve the efficacy of donor lymphocyte infusion (DLI) in a post bone marrow transplantation AML model, and the best effects were seen with a combination of the CpG-A and CpG-B ODN, which appeared to synergize, giving 90% long-term survival when DLI alone gave 0% (222). CpG treatment combined with Flt3 ligand had synergistic antitumor activity in this model. Taken together, these studies demonstrate that CpG-A and CpG-B ODN may have different roles as immunotherapeutic agents depending on the tumor type; CpG-A tends to be most effective against NK-sensitive tumors, and CpG-B is most effective when broader immune activation, including induction of an active immune response, can enhance the efficacy of therapy.

1 Feb 2002

9:54

Annu. Rev. Immunol. 2002.20:709-760. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

732

AR

AR152-22.tex

AR152-22.SGM

LaTeX2e(2001/05/10)

P1: GJC

KRIEG CpG DNA AS AN ADJUVANT FOR CANCER VACCINES The first demonstration of the efficacy of CpG DNA as an adjuvant for a tumor antigen was accidental. Drosophila cell extracts were found to be surprisingly stimulatory to murine B cells, inducing them to express costimulatory molecules and provide bystander costimulation for CD8+ T cells (31). This activity was found to reside in the Drosophila DNA (type-A CpG) and to be lost after treatment with CpG methylase. Drosophila cells transfected to express a tumor antigen in the context of the appropriate class I MHC could induce MHC-compatible spleen cells to respond to a tumor antigen in vitro and to mediate tumor rejection in vivo without any additional signals (31). Both CpG-B and CpG-A ODN are highly effective as vaccine adjuvants, but most cancer vaccine studies have used CpG-B ODN. CpG-B ODN-induced activation of DCs creates a Th1-like cytokine and chemokine environment in the secondary lymphoid organs (142, 223) that promotes cross-priming with strong IFN-γ -secreting cytolytic T cell and antibody responses to peptides and protein antigens independently of T cell help (75, 100, 196, 223–227). The ability of CpGB ODN to induce CTL independently from T cell help is especially noteworthy in light of the failure of DNA vaccines to achieve this (75, 100, 196, 224, 225, 228). In a recent direct comparison of 19 different adjuvants, a CpG-B ODN was the most Th1-like (229). When different adjuvants are combined, the best combinations for inducing Th1 responses are those that contain a CpG-B ODN (230). In a comparison of CpG ODN to complete Freund’s adjuvant (CFA) for immunizing against a lymphoma idiotype (Id), both gave similar high levels of Id-specific antibody (231). The CpG-B ODN was highly effective when used to immunize via the intradermal or subcutaneous routes but was slightly less effective through the intraperitoneal route. Although a dose of 25 µg of the ODN was highly effective, a maximal response was seen at doses of 50–100 µg. Control mice all died within one month of tumor challenge, but mice immunized together with CFA or CpG ODN had 20% or 40% long-term survival, respectively (231). This slightly higher efficacy of the CpG adjuvant compared to CFA may be related to the fact that CpG induced more than twice as much of the IgG 2a anti-Id isotype while CFA induced relatively higher levels of IgG1. The efficacy of CpG DNA was further improved when the Id was conjugated to GMCSF (232). Mice immunized with a GMCSF-Id fusion protein produced high levels of anti-Id antibodies, but almost all of the antibody was IgG1, and only 30% of mice survived a tumor challenge. In contrast, when a CpG ODN was combined with the GMCSFId fusion protein, the antibody level was increased fivefold, was largely IgG2a, and the long-term survival of the mice was improved from 30% to 70% (232). This synergy between CpG and GMCSF is reminiscent of their synergistic DC activation (see above). In immunotherapy of mice bearing a B16 melanoma that expressed ovalbumin (OVA), repeated immunization with OVA protein or peptide together with CpG resulted in a 10–100-fold increase in CTL responses and substantial therapeutic activity even when vaccination was delayed until day 7 (233).

1 Feb 2002

9:54

AR

AR152-22.tex

AR152-22.SGM

LaTeX2e(2001/05/10)

P1: GJC

CPG MOTIFS

733

The antitumor efficacy of adoptive transfer of T cells in an A20 lymphoma model is substantially improved when CpG-B ODN is added to the in vitro culture with APCs and irradiated tumor cells, without apparent induction of autoimmune disease (234). CpG is also an effective adjuvant for a DC tumor vaccine consisting of DC cocultured with irradiated tumor cells, which provide a substantial increase in both prophylactic and therapeutic activity in a murine colon cancer model (235).

Annu. Rev. Immunol. 2002.20:709-760. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

ACTIVATION OF ANTIBODY-DEPENDENT CELLULAR CYTOTOXICITY WITH CpG DNA FOR TUMOR IMMUNOTHERAPY Approximately 70 humanized monoclonal antibodies (mAbs) against tumor antigens are currently in human clinical trials, and three mAbs, Rituximab, Herceptin, and Alemtuzumab, have been approved by the U.S. Food and Drug Administration. Although their mechanisms of action have been controversial, a major antitumor mechanism of mAbs is thought to be antibody-dependent cellular cytotoxicity (ADCC). mAbs specific for tumor cell surface antigens bind to the tumor cell through their antigen binding domains and, through the Fc region of the mAb, bind to cells such as NK cells and neutrophils that express Fc receptors. This binding event is thought to activate the FcR-expressing cell, which results in ADCC. We found that CpG increases ADCC (236), and we therefore hypothesized that the efficacy of antitumor mAbs could be improved by treatment with CpG DNA. We tested this hypothesis in immunocompetent C3H mice with established syngeneic B cell lymphoma, 38C13 (236). Three to five days after IP tumor implantation, the mice were injected with saline, IL-2, or a dose of CpG ODN (given IP) sufficient to cause systemic immune activation and enhanced Fc receptor function, followed by a standard dose of an anti-Id mAb. Although treatment with mAb alone gave only a 10% long-term survival, mice pretreated with CpG ODN had 70%–80% survival (236). Repeated doses of IL-2 only increased survival up to 30%–40%. The CpG ODN was equally effective at promoting survival when administered before the mAb on the same day, or up to two days after the mAb. However, delayed administration of CpG ODN until four days after mAb resulted in survival that was indistinguishable from mAb alone. To further explore the synergistic effect of CpG ODN and mAb therapy, we performed a dose-response evaluation by varying the dose of CpG ODN (237). High doses of CpG ODN (up to 200 µg) enhanced the antitumor activity of mAb more effectively than lower doses of CpG ODN, although some activity was apparent even at doses of 2 µg/mouse. With repeated administration of CpG and mAb, even large tumors could be cured in this model (237). Primary human malignant B cells from patients with various histologies respond to CpG-B ODN in culture, with increased expression of class I and II MHC, CD20, CD40, CD54, CD58, CD80, and CD86, and improved activation of T cells in an allogeneic MLR (238–240). Such changes should make the malignant cells better targets for Rituximab, as well as for immunotherapy in general. Human clinical trials of this form of immunotherapy are under way.

1 Feb 2002

9:54

734

AR

AR152-22.tex

AR152-22.SGM

LaTeX2e(2001/05/10)

P1: GJC

KRIEG

Annu. Rev. Immunol. 2002.20:709-760. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

Immunotherapy of Allergic Diseases with CpG DNA Allergic diseases such as asthma result from Th2-type immune responses against otherwise harmless environmental antigens. Such responses lead to the generation of Th2 T cells, which produce IL-4 and IL-5 and promote the differentiation of B cells into IgE secreting cells. This IgE binds to the high-affinity IgE Fc receptor on the surface of mast cells and basophils. Subsequent exposure of these cells to an allergen results in the binding of the allergen by surface IgE, cross-linking of the IgE Fc receptors, and activation and degranulation of the mast cells or basophils. These cells release a variety of preformed proinflammatory and vasoactive compounds including histamine, prostaglandins, leukotrienes, and cytokines. This results in immediate inflammatory response within 15 min, followed by a secondary late phase reaction several hours later. Almost all current therapeutic efforts against allergic disease have been aimed at the control of the symptoms triggered by mast cell or basophil degranulation. However, a more fundamental approach to disease therapy would be to prevent the initial generation of the Th2-like immune response against the allergen, or to induce a Th1-like response against the allergen. Since Th1 and Th2 immune responses are typically mutually inhibitory, the induction of a Th1-like immune response to an allergen should suppress the Th2-like response. The incidence of asthma and allergic diseases has been increasing dramatically in industrialized nations over the past few decades (reviewed in 241). Although the cause of this atopic epidemic remains unclear, an explanation gaining increasingly wide acceptance has become known as the hygiene hypothesis. Briefly, the hygiene hypothesis posits that the increase in allergy and asthma can be explained by a loss of immune stimulatory exposures during childhood. In the absence of infectious stimulation, the normal development and maturation of the immune system to a Th1-like state fails to occur, and the default Th2 bias of the immature immune system results in atopy (242). Evidence supporting the hygiene hypothesis includes the following observations. First, there is an inverse relationship between atopy and immunization or infection with mycobacteria, measles, and hepatitis A. Second, children with multiple older siblings, who are more likely to develop childhood infections early in life, are less likely to develop atopic disease than only children or firstborns. Third, children cared for in day care settings where they are exposed to large numbers of other children are also protected against atopy. Fourth, children who are treated with antibiotics in the first two years of life have approximately twice the risk of developing allergic disease as those who are not. How could early childhood infections protect against the development of allergic disease? The answer to this question may lie in the potent Th1-like immune stimulatory effects of CpG DNA. Childhood bacterial or viral infections that resulted in exposure to CpG DNA may create a Th1-like environment in the airways. Subsequent allergen exposure would lead to the initiation of a Th1-like immune response against the antigen, which would protect against the subsequent development of allergy. Another mechanism through which CpG DNA may protect against

1 Feb 2002

9:54

AR

AR152-22.tex

AR152-22.SGM

LaTeX2e(2001/05/10)

P1: GJC

Annu. Rev. Immunol. 2002.20:709-760. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

CPG MOTIFS

735

allergic disease is through oral ingestion. It is surprising that oral administration of nuclease-digested DNA to mice causes an approximate 50% drop in serum IgE, IgG1, and IgM levels, which suggests a switch to a Th1-dominant immune milieu (243). Nucleotide supplementation of infant formula improves the response to vaccination compared to human infants fed standard formula (244), but Th1/2 balance was not specifically addressed in this study. It is noteworthy that normal human breast milk contains a significant level of nucleotides (unlike cow milk), the origin of which is unknown. We used a mouse model to test the hypothesis that CpG DNA exposure could prevent the development of allergic disease. Mice were sensitized to a strong Th2like stimulus, schistosome eggs, by IP injection in the presence or absence of a CpG-B ODN. Upon subsequent inhalation challenge with schistosome egg antigen (SEA), control mice developed severe eosinophilic airways disease with high levels of Th2-like cytokines in the airways, eosinophilic infiltrates, and evidence of broncho-constriction (245). In contrast, mice exposed to CpG-B ODN at the time of the initial sensitization were almost completely protected against the development of eosinophilic airways disease. Induction of IgE synthesis was also blocked in such prophylactic experimental models (245, 246). More importantly, CpG DNA could even block eosinophilic airways disease in mice that had already been exposed to the allergen, which suggests possible utility in the treatment of humans with asthma (245). Mice treated with CpG DNA immunotherapy developed a Th1-like immune response to the SEA instead of the Th2-like immune response, but this was not associated with any apparent airways pathology. Our initial hypothesis to explain the therapeutic activity of CpG DNA in this model of asthma was that the CpG DNA was working through induction of expression of the Th1-like cytokines, IL-12 and IFN-γ . However, more recently we have found that even in mice genetically deficient in either or both of these cytokines, CpG DNA can still prevent allergic disease (241). The only difference from wild-type mice was that the Th1-deficient mice required slightly higher doses of the CpG-B ODN in order to prevent disease development. Further studies have confirmed the IL-12 independence of CpG DNA and demonstrated that its immunotherapeutic activity in this model does not depend on NK or B cells (247). Further studies will be required to determine the mechanism of action of CpG DNA in the absence of these classical Th1-like cytokines. CpG-B ODN also induces the production of other Th1-like cytokines, such as IL-18 and IFN-α, which may be sufficient to prevent disease development. CpG DNA can inhibit production of IgE antibodies in human PBMCs stimulated in vitro with IL-4 plus anti-CD40, through a mechanism dependent upon CpG-induced IFN-γ and IL-12 production (248). CpG-B ODN given systemically (IP) or intranasally prevent the sensitization of mice to ovalbumin within one day of CpG administration (249). Since the conventional treatment for allergic airways disease involves the use of corticosteroids, it is noteworthy that a single dose of CpG DNA inhibited airway eosinophilia at least as effectively as daily injections of steroid for seven days (249). Moreover, only CpG therapy redirected the immune response toward a Th1-like response. These

1 Feb 2002

9:54

Annu. Rev. Immunol. 2002.20:709-760. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

736

AR

AR152-22.tex

AR152-22.SGM

LaTeX2e(2001/05/10)

P1: GJC

KRIEG

basic findings have been confirmed by Sur et al., who showed that a single dose of CpG DNA could prevent sensitization to ragweed allergen for six weeks (250). However, in contrast to the finding of Kline et al. that IFN-γ knockout mice were still protected against eosinophilic airways inflammation by CpG DNA, Sur et al. found that administration of CpG ODN to IFN-γ –deficient mice did not protect against airway eosinophilia. The difference between these two studies may be that the ODN used by the latter authors contained suboptimal CpG motifs and/or the dose response to treatment, which was not determined. CpG ODN also inhibit the induction of Th2-like immune responses to cedar pollen (251) and birch pollen allergens in mice (252, 253). To test the ability of CpG ODN to reverse Th2 pathology in mice after repeated airway allergen challenges, Serebrisky et al. sensitized AKR mice to conalbumin and then gave them two airway challenges at one week intervals. Administration of a CpG ODN 24 h after the first allergen challenge significantly reduced, but did not eliminate, the airway inflammatory response (254). Although in general CpG treatment induces expression of B7.2, as noted above, in this experimental system CpG treatment was associated with reduced B7.2 mRNA expression in the lung. It is intriguing that the addition of CpG-B ODN to human PBMCs from allergic donors results in decreased in vitro production of IgE, which suggests potential clinical utility for CpG DNA in the immunotherapy of allergic diseases (112).

TOXICOLOGY AND ADVERSE EFFECTS OF CpG DNA Effects of CpG DNA on Immune Tolerance and Autoimmunity TOLERANCE TO DNA Several murine experimental models have been established in which anti-DNA Ab production can be elicited. It is very difficult to induce production of antibodies against native, unmodified DNA, and studies have firmly established that dsDNA is an extremely poor antigen (reviewed in 255, 256). Mice transgenic for anti-ssDNA and anti-dsDNA VH genes have been useful to investigate how tolerance influences the ability of B cells to produce anti-ssDNA and antidsDNA Ab. In these transgenic mice, B cell precursors for anti-ssDNA–producing cells mature but are anergic and do not secrete Ab (257). In contrast, precursors for anti-dsDNA B cells are normally deleted in the bone marrow (central deletion) (258) or undergo receptor editing, a process in which their maturation is arrested, and they may have continued light chain gene rearrangement until they produce a surface Ig that does not bind self-antigen (259). Anti-dsDNA B cells appear to be incompletely deleted in mice genetically predisposed to autoimmunity (260). Even if central deletion of anti-dsDNA B cells functions perfectly, anti-dsDNAspecific B cells can arise in the periphery through somatic mutation and affinity maturation during an immune response and are normally eliminated, presumably through apoptosis (261). Thus, tolerance mechanisms of anergy, receptor editing, and deletion normally may prevent the production of anti-ssDNA and dsDNA Ab, respectively.

1 Feb 2002

9:54

AR

AR152-22.tex

AR152-22.SGM

LaTeX2e(2001/05/10)

P1: GJC

Annu. Rev. Immunol. 2002.20:709-760. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

CPG MOTIFS

737

These transgenic studies also provide an explanation for the observation that nonautoimmune mice treated with polyclonal B cell mitogens produce anti-ssDNA Ab but not anti-dsDNA Ab (reviewed in 255). Polyclonal mitogens appear able to partially overcome peripheral anergy, thereby activating Ab production in preexisting low affinity anti-ssDNA–specific B cells. However, a polyclonal mitogen would not induce production of anti-dsDNA Ab because development of the precursor B cells would have been blocked in the bone marrow. The production of pathogenic anti-DNA Ab in lupus-prone mice appears to require the loss of T cell tolerance to DNA binding proteins in chromatin (reviewed in 262). One way in which autoreactive T cells can provide help to DNA-specific B cells is through intermolecular help. The first step in this model is for a B cell specific for DNA to take up a DNA-protein complex and be activated either by CpG motifs in the DNA or by other stimuli to express costimulatory signals. The activated B cell then presents the protein to a specific T cell, which is then activated to provide further B cell help for isotype switching. Thus, T cells specific for foreign DNA binding proteins are activated and provide help to DNA-specific B cells, driving anti-dsDNA Ab production. For example, mice immunized with a 27–amino acid DNA-binding peptide from a Trypanosoma protein together with native mammalian DNA in Freund’s adjuvant produce high levels of anti-dsDNA Ab that are similar in isotype, V region usage, and specificity to the spontaneous Ab produced by lupus-prone mice (263, 264). Another model system in which immunization with a single foreign DNA binding protein induces anti-dsDNA Ab production involves injecting a DNA vaccine encoding either the SV40 or BK virus T antigens under the control of the CMV promoter (265). In this case, the APC activation stimulus presumably comes from the CpG motifs in the plasmid instead of from CFA. Epitope spreading appears to occur in this system, mimicking human autoimmunity (266). Mice injected with plasmids encoding non-DNA binding proteins did not produce anti-dsDNA Ab in these studies, which demonstrates that immune activation by the plasmid CpG is insufficient to break self-tolerance to DNA. Normal mice immunized with bacterial E. coli (EC) DNA (or bDNA) complexed with methylated BSA in CFA produce anti-dsDNA antibodies specific for EC DNA, [but not to vertebrate (calf thymus, CT) DNA], but mice immunized with CT DNA in the same manner do not make any anti-dsDNA Ab (267). The protein complexes and CFA should provide T cell help, as was required in the models described in the preceding paragraph. In contrast to their results in normal mice, immunization of preautoimmune NZB/NZW mice led to the production of antibodies to both EC and CT DNA (268). However, the SLE-prone mice immunized with bDNA had a decreased severity of nephritis compared to unimmunized mice, underscoring the fact that anti-dsDNA Ab production does not always correlate with disease (269). Thus, administration of CpG DNA to autoimmune-prone mice in these model systems does not aggravate disease but rather attenuates it. This failure of CpG treatment to flare autoimmune disease has also been demonstrated with administration of naked CpG DNA. For example, Mor et al. administered four injections of CpG DNA in the form of a DNA vaccine to lupus-prone

1 Feb 2002

9:54

738

AR

AR152-22.tex

AR152-22.SGM

LaTeX2e(2001/05/10)

P1: GJC

KRIEG

Annu. Rev. Immunol. 2002.20:709-760. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

NZB X NZW F1 mice and evaluated the mice for B cells secreting various autoantibodies (270). Although there was a modest increase in the number of B cells secreting IgG anti-DNA antibodies, there was no apparent glomerulonephritis or autoimmune disease (270). In summary, even repeated CpG-augmented immunization caused no clinically apparent adverse consequences in either normal or lupus-prone mice. TOLERANCE TO CNS ANTIGENS Experimental autoimmune encephalomyelitis (EAE) models have some similarities to human multiple sclerosis (MS). In the most common of these models, mice or rats of particular sensitive strains are immunized with either of two abundant CNS proteins, proteolipid protein (PLP) or myelin basic protein (MBP), together with CFA. This leads to activation of autoreactive T cells that can cause severe CNS disease and even death. However, disease is usually transient and may not develop into the relapsing-remitting pattern typical of human MS. Recent studies suggest that EAE may not result from breaking immune tolerance to self-antigens. The SJL mouse strain commonly used is highly unusual in that it does not express the dominant epitope of PLP in the thymus (271, 272). Thus, these mice are ignorant, not truly tolerant to this self-antigen, which potentially explains why induction of immune responses to the antigen is comparatively easy in this mouse strain (271, 272). It is of interest that injection of bDNA or CpG ODN into the CNS of non-EAE-prone mice or rats does not induce anything more than a transient inflammatory meningitis (123, 124). CpG DNA induces high levels of IL-12 expression, which previously has been linked to the pathogenesis of EAE. Like IL-12, CpG DNA can induce EAE in certain experimental models. For example, if lymph node cells from mice immunized with CFA and MBP are cultured in vitro with a CpG-B ODN, the IL-12 activates autoreactive Th1 T cells, which cause EAE upon adoptive transfer into susceptible mice (273). CpG can also be used in vivo to replace CFA in immunizing mice or rats against CNS antigens, resulting in EAE (274, 274a). The EAE that occurred was transient and followed by spontaneous recovery. Mice or rats injected with a CpG-B ODN without CNS antigens do not develop EAE. Moreover, CpG does not induce a relapsing pattern of EAE, nor can it induce EAE in rats not genetically susceptible (274a). On the other hand, CpG DNA treatment may aggravate induction of chronic demyelinating EAE induced by immunization with PLP in CFA. Tsunoda et al. showed that pretreatment of susceptible SJL mice with 100 µg of a DNA plasmid (CpG-A type) weekly for one to three weeks worsened the disease (275). This model requires injection of the PLP with CFA. It is unclear whether a CpG-B ODN would have the same aggravating effect as the DNA plasmid. Depending on the timing and dosage of administration, CpG DNA can also decrease the severity of EAE, actually protecting against disease. Ruiz et al. have reported that administration of 100 µg of DNA vaccines (given in 2 doses one week apart) encoding PLP protects SJL mice against the acute phase of EAE

1 Feb 2002

9:54

AR

AR152-22.tex

AR152-22.SGM

LaTeX2e(2001/05/10)

P1: GJC

Annu. Rev. Immunol. 2002.20:709-760. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

CPG MOTIFS

739

induction ten days later (276). These investigators reported that control plasmid DNA lacking the PLP had no positive or negative effect on disease. Lobell et al. used a DNA vaccine encoding MBP to suppress EAE in Lewis rats when this was induced three to five weeks after vaccination (277). In further studies, Lobell et al. showed that the protective effect of DNA vaccination was significantly enhanced if additional CpG motifs were cloned into the vector, and reduced if the vector was depleted of CpG motifs, which demonstrates an important beneficial effect of CpG for preventing autoimmunity (278). The mechanism of the protective effect of CpG motifs against EAE has been further investigated by Boccacio et al., who demonstrated in the same Lewis rat model that it is the ability of CpG motifs to enhance IFN-γ production that accounts for its protective function (279). In fact, repeated injection of mice with 50 µg of a CpG-B ODN starting 4 days prior to EAE induction led to a marked reduction in the subsequent disease severity, compared to mice treated with a control GpC ODN (279). In all of these studies, immunization resulted in increased immune responses to the CNS antigen, but the altered cytokine profile of the induced T cells appears to reduce, rather than increase, the risk of autoimmune disease. A novel DNA vaccine-based approach to EAE immunotherapy is based on DNA vaccination against Fas ligand (FasL) to induce production of autoantibodies against FasL, which then protected against EAE induction in susceptible mice (280). A similar strategy was successful at inducing autoimmunity against specific C-C chemokines, which are apparently required for mediating EAE pathogenesis (281). The efficacy of DNA vaccination to self-cytokines and chemokines is not limited to the prevention of EAE, but it can also protect against adjuvantinduced arthritis in susceptible Lewis rats (282). These studies indicate that DNA vaccination against self-antigens may break self-tolerance yet still protect against autoimmune disease development. Infection of mice with Theiler’s murine encephalomyelitis virus is another model that represents some features of human MS. Administration of DNA vaccines (which contain CpG motifs) to mice infected with Theiler’s murine encephalomyelitis virus causes exacerbation of the demyelinating disease in a dosedependent manner (275). Interpretation of these results is not simple because the effects may be explained by the induction of proinflammatory cytokines, and loss of immune tolerance in response to CpG was not demonstrated. As these toxic manifestations have not been seen when CpG DNA is administered to mice systemically in the absence of this viral infection, and because this virus does not infect humans, the relevance for the therapeutic use of CpG DNA in humans is unclear. In summary, these studies show that immunization of rodents with high doses of purified CNS antigens formulated with CpG ODN leads to the induction of acute transient EAE with similar efficiency to immunization with CFA. However, treatment of rodents with DNA vaccines or CpG-B ODN prior to immunization with high-dose CNS antigens generally protects against the acute phase of disease development.

1 Feb 2002

9:54

740

AR

AR152-22.tex

AR152-22.SGM

LaTeX2e(2001/05/10)

P1: GJC

KRIEG

Annu. Rev. Immunol. 2002.20:709-760. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

TOLERANCE TO CARDIAC ANTIGENS If mice are immunized with foreign antigens that are molecular mimics of self-antigens, then it is possible to induce an autoimmune response. This disease-inducing mechanism has been demonstrated in a report where mice immunized with CpG DNA plus chlamydia-derived antigens that are molecular mimics of cardiac antigens developed an autoimmune myocarditis (283). This report demonstrates that CpG DNA can act as an adjuvant for an autoimmune response to a cardiac antigen, apparently by the mechanism of molecular mimicry. TOLERANCE TO COLLAGEN Immunization of certain mouse or rat strains with bovine collagen in the presence of a strong Th1-like adjuvant induces an autoimmune arthritis, perhaps through a molecular mimicry mechanism similar to that described in the preceding paragraph. If the mice are first primed with injection of CpG DNA to create an enhanced in vivo Th1 response, then they develop a more severe inflammatory arthritis in response to bovine collagen immunization (284). Together with the previous report of cardiac autoimmunity, this finding supports the conclusion that CpG DNA may enhance the ability of molecular mimics of self-antigens to induce autoimmunity. TOLERANCE TO MHC MOLECULES IN TRANSGENIC MICE B cells with antigen receptors that are specific for self–class I MHC molecules are normally deleted in the bone marrow. However, if the class I MHC molecule is expressed as a transgene under the control of a liver-specific promoter (i.e., not expressed in the bone marrow), the B cells are able to mature to leave the bone marrow and then undergo peripheral deletion in the liver (285). Administration of CpG DNA to these transgenic mice fails to rescue the B cells against deletion, even though administration of a phage expressing the self-antigen can rescue the autoreactive B cells (285). This demonstrates that immune activation by CpG DNA is unable to overcome self-tolerance in a model where it can be overcome by antigen without CpG. TOLERANCE TO HEPATITIS B SURFACE ANTIGEN (HBsAg) IN TRANSGENIC MICE Mice transgenic for most of the HBV genome (core gene deleted) express viral genes such as HBsAg under the control of the endogenous HBV promoter, principally in the liver (286). These mice have high circulating levels of HBsAg but no accumulation in the liver and no pathology. B and T cell tolerance in these Tg mice can be overcome by immunization with recombinant HBsAg in CFA or with DNA vaccination (286). It is surprising that the resulting immune response clears circulating HBsAg and markedly reduces HBsAg mRNA expression in the liver without causing a cytopathic effect (286). Adoptive transfer experiments showed that either CD4 or CD8 T cells could cause the noncytolytic control of viral expression, and that control required IFN-γ , suggesting a Th1 type of response (287). Immunization of these Tg mice with CpG ODN together with HBsAg also induced a noncytolytic T cell–dependent control of viral expression but to a lesser extent than seen after DNA vaccination (287a). Even though T cell tolerance in these mice was abrogated, no apparent autoimmune disease occurred. Moreover, administration of CpG ODN

1 Feb 2002

9:54

AR

AR152-22.tex

AR152-22.SGM

LaTeX2e(2001/05/10)

P1: GJC

CPG MOTIFS

741

alone without HBsAg caused no loss of self-tolerance at all. These studies again demonstrate that injection of CpG DNA alone does not overcome self-tolerance and that even when it is administered with a self-antigen under conditions that can overcome self-tolerance, disease need not result.

Annu. Rev. Immunol. 2002.20:709-760. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

Induction of Nonautoimmune Inflammatory Pathology by CpG DNA Under some experimental conditions, CpG DNA can trigger immune-mediated inflammation that is not autoimmune. For example, when injected into joints, CpG DNA induces a transient inflammatory arthritis with a peak in severity at 3 days, followed by gradual waning (288, 289). In these reports, the proinflammatory effects of CpG DNA injected into the joint appear to result in a self-limited local inflammation, but induction of antigen-specific autoimmunity was not observed. In fact, the arthritis did not depend at all on T or B cells but only on macrophages, which were activated by the CpG DNA to make proinflammatory cytokines in the joint (290). Furthermore, there were no effects on the joint if the DNA was injected systemically instead of directly into the joint (288). TNFα-deficient mice develop a milder disease, which suggests an important pathogenic role for this cytokine (290). The arthritis was not progressive, even if the CpG DNA was injected repeatedly, and cartilage or bone destruction, which are characteristic of autoimmune arthritis, did not develop. Thus, the relevance to human destructive autoimmune arthritis is uncertain, although it is intriguing that bacterial DNA has been described within synovial tissues of some arthritis patients (291). In a like manner, CpG-B ODN injected into the CNS activates macrophage-like cells such as glial cells and astrocytes, which results in a transient inflammation lasting approximately five days, but no apparent permanent harm (123, 124).

CpG DNA as A Trigger for the Systemic Inflammatory Response Syndrome Pretreatment with CpG DNA protects against infection or tumor challenge and prevents sensitization to allergens, but it can also be lethal in the wrong setting, such as in mice that have been primed with endotoxin (98) or D-galactosamine (186, 292) by inducing the systemic inflammatory response syndrome (SIRS). LPS and bDNA or CpG-B ODN synergize for inflammatory cytokine production (98) probably through several mechanisms (128, 129). On the other hand, pretreatment with CpG antagonizes the toxic effects of LPS inhalation through an IL-10-mediated mechanism (293).

A Summary of Past Human Therapy with CpG DNA Natural environmental exposures to CpG ODN in the form of infections are quite frequent and, although viral infections may induce anti-dsDNA antibody production (294), chronic infections do not lead to an increased risk of autoimmune disease in humans. A long history of human therapy with crude forms of CpG

1 Feb 2002

9:54

Annu. Rev. Immunol. 2002.20:709-760. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

742

AR

AR152-22.tex

AR152-22.SGM

LaTeX2e(2001/05/10)

P1: GJC

KRIEG

DNA supports its safety. It appears likely that the immune stimulatory effects of CFA and BCG can be explained, at least in part, by their CpG DNA content (reviewed in 10). These immune modulators have been administered to tens of millions of human subjects with resulting local inflammatory responses and some increased levels of autoantibodies but without induction of autoimmune disease. Tokunaga and colleagues demonstrated in 1984 that bacterial DNA was responsible for the immune stimulatory effects of BCG; they then went on to perform several human clinical trials using BCG DNA without noting any induction of autoimmune diseases (reviewed in 10). More recently, many humans have received DNA vaccines in clinical trials, and there were no reports of autoantibody induction or autoimmune disease. Patients who have received nonviral vectors that contained CpG DNA in gene therapy clinical trials have shown signs of immune activation with flu-like symptoms but have not developed autoimmune diseases (reviewed in 295, 296). These data suggest that administration of PO DNA containing CpG motifs to humans induces immune activation but not autoimmune disease. Of course, it is possible that PS ODN with CpG motifs may have different effects on immune tolerance from those of native DNA. However, here too there are published data that indicate administration of CpG ODN to humans appears safe. By chance, several of the approximately 20 antisense ODN that have entered human clinical trials have had CpG motifs that could stimulate human B cell and DC expression of costimulatory molecules. Hundreds of humans have received these antisense ODN with no reports to date of association with autoimmunity and no evidence of anti-DNA antibody formation (297). These studies indicate that CpG DNA does not generally abrogate B or T cell tolerance or induce autoantibody production or autoimmune disease, even in genetically predisposed individuals. ACKNOWLEDGMENT The author thanks Denise Arsenault for assistance with manuscript preparation. Visit the Annual Reviews home page at www.annualreviews.org

LITERATURE CITED 1. Dempsey PW, Allison ME, Akkaraju S, Goodnow CC, Fearon DT. 1996. C3d of complement as a molecular adjuvant: bridging innate and acquired immunity. Science 271(5247):348–50 2. Kumar A, Yang YL, Flati V, Der S, Kadereit S, Deb A, Haque J, Reis L, Weissmann C, Williams BR. 1997. Deficient cytokine signaling in mouse embryo fibroblasts with a targeted deletion in the

PKR gene: role of IRF-1 and NF-kappaB. EMBO J. 16(2):406–16 3. Bird AP. 1987. CpG islands as gene markers in the vertebrate nucleus. Trends Genet. 3:342–47 4. Krieg AM, Yi AK, Matson S, Waldschmidt TJ, Bishop GA, Teasdale R, Koretzky GA, Klinman DM. 1995. CpG motifs in bacterial DNA trigger direct B-cell activation. Nature 374(6522):6546–49

1 Feb 2002

9:54

AR

AR152-22.tex

AR152-22.SGM

LaTeX2e(2001/05/10)

P1: GJC

Annu. Rev. Immunol. 2002.20:709-760. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

CPG MOTIFS 5. Medzhitov R, Janeway CA Jr. 1997. Innate immunity: impact on the adaptive immune response. Curr. Opin. Immunol. 9(1):4–9 6. Wiemann B, Starnes CO. 1994. Coley’s toxins, tumor necrosis factor and cancer research: a historical perspective. Pharmacol. Ther. 64(3):529–64 7. Morales A. 1978. Adjuvant immunotherapy in superficial bladder cancer. Natl. Cancer Inst. Monogr. 49:315–19 8. Tokunaga T, Yamamoto H, Shimada S, Abe H, Fukuda T, Fujisawa Y, Furutani Y, Yano O, Kataoka T, Sudo T. 1984. Antitumor activity of deoxyribonucleic acid fraction from Mycobacterium bovis BCG. I. Isolation, physicochemical characterization, and antitumor activity. J. Natl. Cancer Inst. 72(4):955– 62 9. Yamamoto S, Kuramoto E, Shimada S, Tokunaga T. 1988. In vitro augmentation of natural killer cell activity and production of interferon-α/β and -γ with deoxyribonucleic acid fraction from Mycobacterium bovis BCG. Jpn. J. Cancer Res. 79(7):866–73 10. Tokunaga T, Yamamoto T, Yamamoto S. 1999. How BCG led to the discovery of immunostimulatory DNA. Jpn. J. Infect. Dis. 52(1):1–11 11. Yamamoto S, Yamamoto T, Shimada S, Kuramoto E, Yano O, Kataoka T, Tokunaga T. 1992. DNA from bacteria, but not from vertebrates, induces interferons, activates natural killer cells and inhibits tumor growth. Microbiol. Immunol. 36(9):983–97 12. Kuramoto E, Yano O, Kimura Y, Baba M, Makino T, Yamamoto S, Yamamoto T, Kataoka T, Tokunaga T. 1992. Oligonucleotide sequences required for natural killer cell activation. Jpn. J. Cancer Res. 83(11):1128–31 13. Tokunaga T, Yamamoto S, Namba K. 1988. A synthetic single-stranded DNA, poly(dG,dC), induces interferon-α/β and -γ , augments natural killer activity, and

14.

15.

16.

17.

18.

19.

20.

21.

22.

743

suppresses tumor growth. Jpn. J. Cancer Res. 79(6):682–86 Messina JP, Gilkeson GS, Pisetsky DS. 1991. Stimulation of in vitro murine lymphocyte proliferation by bacterial DNA. J. Immunol. 147(6):1759–64 Messina JP, Gilkeson GS, Pisetsky DS. 1993. The influence of DNA structure on the in vitro stimulation of murine lymphocytes by natural and synthetic polynucleotide antigens. Cell Immunol. 147(1):148–57 Pisetsky DS, Reich CF III. 1998. The influence of base sequence on the immunological properties of defined oligonucleotides. Immunopharmacology 40(3):199–208 Suzuki K, Mori A, Ishii KJ, Saito J, Singer DS, Klinman DM, Krause PR, Kohn LD. 1999. Activation of target-tissue immune-recognition molecules by double-stranded polynucleotides. Proc. Natl. Acad. Sci. USA 96(5):2285–90 Ishii KJ, Suzuki K, Coban C, Takeshita F, Itoh Y, Matoba H, Kohn LD, Klinman DM. 2001. Genomic DNA released by dying cells induces the maturation of APCs. J. Immunol. 167(5):2602–7 Tanaka T, Chu CC, Paul WE. 1992. An antisense oligonucleotide complementary to a sequence in Ig 2b increases γ 2b germline transcripts, stimulates B cell DNA synthesis, and inhibits immunoglobulin secretion. J. Exp. Med. 175(2):597–607 Branda RF, Moore AL, Mathews L, McCormack JJ, Zon G. 1993. Immune stimulation by an antisense oligomer complementary to the rev gene of HIV-1. Biochem. Pharmacol. 45(10):2037–43 Pisetsky DS, Reich CF. 1993. Stimulation of murine lymphocyte proliferation by a phosphorothioate oligonucleotide with antisense activity for herpes simplex virus. Life Sci. 54:101–7 McIntyre KW, Lombard-Gillooly K, Perez JR, Kunsch C, Sarmiento UM, Larigan JD, Landreth KT, Narayanan R.

1 Feb 2002

9:54

744

23.

Annu. Rev. Immunol. 2002.20:709-760. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

24.

25.

26.

27.

28.

29.

30.

AR

AR152-22.tex

AR152-22.SGM

LaTeX2e(2001/05/10)

P1: GJC

KRIEG 1993. A sense phosphorothioate oligonucleotide directed to the initiation codon of transcription factor NF-κB p65 causes sequence-specific immune stimulation. Antisense Res. Dev. 3(4):309–22 Krieg AM, Steinberg AD. 1990. Analysis of thymic endogenous retroviral expression in murine lupus. Genetic and immune studies. J. Clin. Invest. 86(3):809–16 Krieg AM, Gause WC, Gourley MF, Steinberg AD. 1989. A role for endogenous retroviral sequences in the regulation of lymphocyte activation. J. Immunol. 143(8):2448–51 Krieg AM, Tonkinson J, Matson S, Zhao Q, Saxon M, Zhang LM, Bhanja U, Yakubov L, Stein CA. 1993. Modification of antisense phosphodiester oligodeoxynucleotides by a 50 cholesteryl moiety increases cellular association and improves efficacy. Proc. Natl. Acad. Sci. USA 90(3):1048–52 Mojcik CF, Gourley MF, Klinman DM, Krieg AM, Gmelig-Meyling F, Steinberg AD. 1993. Administration of a phosphorothioate oligonucleotide antisense to murine endogenous retroviral MCF env causes immune effects in vivo in a sequence-specific manner. Clin. Immunol. Immunopathol. 67(2):130–36 Krieg AM. 1998. Leukocyte stimulation by oligodeoxynucleotides. In Applied Antisense Oligonucleotide Technology, ed. CA Stein, AM Krieg, pp. 431–48. New York: Wiley Krieg AM, Wu T, Weeratna R, Efler SM, Love-Homan L, Yang L, Yi AK, Short D, Davis HL. 1998. Sequence motifs in adenoviral DNA block immune activation by stimulatory CpG motifs. Proc. Natl. Acad. Sci. USA 95(21):12,631–36 Han J, Zhu Z, Hsu C, Finley WH. 1994. Selection of antisense oligonucleotides on the basis of genomic frequency of the target sequence. Antisense Res. Dev. 4(1):53–65 Brown WC, Estes DM, Chantler SE, Kegerreis KA, Suarez CE. 1998. DNA

31.

32.

33.

34.

35.

36.

37.

38.

39.

and a CpG oligonucleotide derived from Babesia bovis are mitogenic for bovine B cells. Infect. Immun. 66(11):5423–32 Sun S, Cai Z, Langlade-Demoyen P, Kosaka H, Brunmark A, Jackson MR, Peterson PA, Sprent J. 1996. Dual function of Drosophila cells as APCs for naive CD8+ T cells: implications for tumor immunotherapy. Immunity 4(6):555– 64 Sun S, Beard C, Jaenisch R, Jones P, Sprent J. 1997. Mitogenicity of DNA from different organisms for murine B cells. J. Immunol. 159(7):3119–25 Karlin S, Ladunga I, Blaisdell BE. 1994. Heterogeneity of genomes: measures and values. Proc. Natl. Acad. Sci. USA 91(26): 12,837–41 Shpaer EG, Mullins JI. 1990. Selection against CpG dinucleotides in lentiviral genes: a possible role of methylation in regulation of viral expression. Nucleic Acids Res. 18(19):5793–97 Krieg AM. 1996. Lymphocyte activation by CpG dinucleotide motifs in prokaryotic DNA. Trends Microbiol. 4(2):73–76 Goto M, Washio T, Tomita M. 2000. Causal analysis of CpG suppression in the Mycoplasma genome. Microb. Comp. Genomics 5(1):51–58 Yi AK, Chang M, Peckham DW, Krieg AM, Ashman RF. 1998. CpG oligodeoxyribonucleotides rescue mature spleen B cells from spontaneous apoptosis and promote cell cycle entry. J. Immunol. 160(12):5898–906 [Erratum J. Immunol. 1999, Jul. 15. 163 (2):1093] Rankin R, Pontarollo R, Iannou X, Krieg AM, Hecker R, Babiuk LA, van Drunen Littel-van den Hurk S. 2001. CpG motif identification for veterinary and laboratory species demonstrates that sequence recognition is highly conserved. Antisense Nucleic Acid Drug Dev. 11:333–40 Hartmann G, Krieg AM. 2000. Mechanism and function of a newly identified CpG DNA motif in human primary B cells. J. Immunol. 164(2):944–53

1 Feb 2002

9:54

AR

AR152-22.tex

AR152-22.SGM

LaTeX2e(2001/05/10)

P1: GJC

Annu. Rev. Immunol. 2002.20:709-760. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

CPG MOTIFS 40. Iho S, Yamamoto T, Takahashi T, Yamamoto S. 1999. Oligodeoxynucleotides containing palindrome sequences with internal 50 -CpG-30 act directly on human NK and activated T cells to induce IFN-γ production in vitro. J. Immunol. 163(7):3642–52 41. Kanellos TS, Sylvester ID, Butler VL, Ambali AG, Partidos CD, Hamblin AS, Russell PH. 1999. Mammalian granulocyte-macrophage colony-stimulating factor and some CpG motifs have an effect on the immunogenicity of DNA and subunit vaccines in fish. Immunology 96(4):507– 10 42. Hartmann G, Weeratna RD, Ballas ZK, Payette P, Blackwell S, Suparto I, Rasmussen WL, Waldschmidt M, Sajuthi D, Purcell H, Davis HL, Krieg AM. 2000. Delineation of a CpG phosphorothioate oligodeoxynucleotide for activating primate immune responses in vitro and in vivo. J. Immunol. 164(3):1617–24 43. Ballas ZK, Rasmussen WL, Krieg AM. 1996. Induction of NK activity in murine and human cells by CpG motifs in oligodeoxynucleotides and bacterial DNA. J. Immunol. 157(5):1840–45 44. Yamamoto T, Yamamoto S, Kataoka T, Tokunaga T. 1994. Lipofection of synthetic oligodeoxyribonucleotide having a palindromic sequence of AACGTT to murine splenocytes enhances interferon production and natural killer activity. Microbiol. Immunol. 38(10):831–36 45. Zhao Q, Matson S, Herrera CJ, Fisher E, Yu H, Krieg AM. 1993. Comparison of cellular binding and uptake of antisense phosphodiester, phosphorothioate, and mixed phosphorothioate and methylphosphonate oligonucleotides. Antisense Res. Dev. 3(1):53–66 46. Matson S, Krieg AM. 1992. Nonspecific suppression of [3H]thymidine incorporation by “control” oligonucleotides. Antisense Res. Dev. 2(4):325–30 47. Kenter AL, Tredup J. 1991. High expression of a 30 –50 exonuclease activity is spe-

48.

49.

50.

51.

52.

53.

54.

55.

56.

57.

745

cific to B lymphocytes. Mol. Cell Biol. 11(9):4398–404 Liang H, Nishioka Y, Reich CF, Pisetsky DS, Lipsky PE. 1996. Activation of human B cells by phosphorothioate oligodeoxynucleotides. J. Clin. Invest. 98(5):1119–29 Hartmann G, Weiner GJ, Krieg AM. 1999. CpG DNA: a potent signal for growth, activation, and maturation of human dendritic cells. Proc. Natl. Acad. Sci. USA 96(16):9305–10 Stein CA, Subasinghe C, Shinozuka K, Cohen JS. 1988. Physicochemical properties of phosphorothioate oligodeoxynucleotides. Nucleic Acids Res. 16(8):3209– 21 Stein CA, Cheng YC. 1993. Antisense oligonucleotides as therapeutic agents– is the bullet really magical? Science 261(5124):1004–12 Stein CA, Krieg AM. 1994. Problems in interpretation of data derived from in vitro and in vivo use of antisense oligodeoxynucleotides [editorial]. Antisense Res. Dev. 4(2):67–69 Zhao Q, Waldschmidt T, Fisher E, Herrera CJ, Krieg AM. 1994. Stage-specific oligonucleotide uptake in murine bone marrow B-cell precursors. Blood 84(11): 3660–66 Crooke RM. 1998. In vitro cellular uptake, distribution, and metabolism of oligonucleotides. In Antisense Research and Application, ed. ST Crooke, pp. 103–40. Berlin: Springer Perez JR, Li Y, Stein CA, Majumder S, van Oorschot A, Narayanan R. 1994. Sequence-independent induction of Sp1 transcription factor activity by phosphorothioate oligodeoxynucleotides. Proc. Natl. Acad. Sci. USA 91(13):5957–61 Wang XZ, Ron D. 1996. Stress-induced phosphorylation and activation of the transcription factor CHOP (GADD153) by p38 MAP kinase. Science 272(5266):1347–49 Burgess TL, Fisher EF, Ross SL, Bready

1 Feb 2002

9:54

746

Annu. Rev. Immunol. 2002.20:709-760. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

58.

59.

60.

61.

62.

63.

64.

AR

AR152-22.tex

AR152-22.SGM

LaTeX2e(2001/05/10)

P1: GJC

KRIEG JV, Qian YX, Bayewitch LA, Cohen AM, Herrera CJ, Hu SS, Kramer TB. 1995. The antiproliferative activity of c-myb and c-myc antisense oligonucleotides in smooth muscle cells is caused by a nonantisense mechanism. Proc. Natl. Acad. Sci. USA 92(9):4051–55 Fennewald SM, Rando RF. 1995. Inhibition of high-affinity basic fibroblast growth factor binding by oligonucleotides. J. Biol. Chem. 270(37):21,718– 21 Guvakova MA, Yakubov LA, Vlodavsky I, Tonkinson JL, Stein CA. 1995. Phosphorothioate oligodeoxynucleotides bind to basic fibroblast growth factor, inhibit its binding to cell surface receptors, and remove it from low affinity binding sites on extracellular matrix. J. Biol. Chem. 270(6):2620–27 Kitajima I, Unoki K, Maruyama I. 1999. Phosphorothioate oligodeoxynucleotides inhibit basic fibroblast growth factorinduced angiogenesis in vitro and in vivo. Antisense Nucleic Acid Drug Dev. 9(2):233–39 Brown DA, Kang SH, Gryaznov SM, DeDionisio L, Heidenreich O, Sullivan S, Xu X, Nerenberg MI. 1994. Effect of phosphorothioate modification of oligodeoxynucleotides on specific protein binding. J. Biol. Chem. 269(43):26,801–5 Khaled Z, Benimetskaya L, Zeltser R, Khan T, Sharma HW, Narayanan R, Stein CA. 1996. Multiple mechanisms may contribute to the cellular anti-adhesive effects of phosphorothioate oligodeoxynucleotides. Nucleic Acids Res. 24(4):737– 45 Hartmann G, Krug A, Waller-Fontaine K, Endres S. 1996. Oligodeoxynucleotides enhance lipopolysaccharide-stimulated synthesis of tumor necrosis factor: dependence on phosphorothioate modification and reversal by heparin. Mol. Med. 2(4):429–38 Monteith DK, Henry SP, Howard RB, Flournoy S, Levin AA, Bennett CF,

65.

66.

67.

68.

69.

70.

71.

72.

Crooke ST. 1997. Immune stimulation– a class effect of phosphorothioate oligodeoxynucleotides in rodents. Anticancer Drug Des. 12(5):421–32 Tam RC, Wu-Pong S, Pai B, Lim C, Chan A, Thomas DF, Milovanovic T, Bard J, Middleton PJ. 1999. Increased potency of an aptameric G-rich oligonucleotide is associated with novel functional properties of phosphorothioate linkages. Antisense Nucleic Acid Drug Dev. 9(3):289–300 Henry S, Stecker K, Brooks D, Monteith D, Conklin B, Bennett CF. 2000. Chemically modified oligonucleotides exhibit decreased immune stimulation in mice. J. Pharmacol. Exp. Ther. 292(20):468–79 Zhao Q, Temsamani J, Iadarola PL, Jiang Z, Agrawal S. 1996. Effect of different chemically modified oligodeoxynucleotides on immune stimulation. Biochem. Pharmacol. 51(2):173–82 Krieg AM, Matson S, Fisher E. 1996. Oligodeoxynucleotide modifications determine the magnitude of B cell stimulation by CpG motifs. Antisense Nucleic Acid Drug Dev. 6(2):133–39 Sester DP, Beasley SJ, Sweet MJ, Fowles LF, Cronau SL, Stacey KJ, Hume DA. 1999. Bacterial/CpG DNA down-modulates colony stimulating factor-1 receptor surface expression on murine bone marrow-derived macrophages with concomitant growth arrest and factor-independent survival. J. Immunol. 163(12): 6541–50 Boggs RT, McGraw K, Condon T, Flournoy S, Villiet P, Bennett CF, Monia BP. 1997. Characterization and modulation of immune stimulation by modified oligonucleotides. Antisense Nucleic Acid Drug Dev. 7(5):461–71 Tang J, Roskey A, Li Y, Agrawal S. 1995. Enzymatic synthesis of stereoregular (all Rp) oligonucleotide phosphorothioate and its properties. Nucleosides & Nucleotides 14(3–5):985–90 Yu D, Kandimalla ER, Roskey A, Zhao Q, Chen L, Chen J, Agrawal S. 2000.

1 Feb 2002

9:54

AR

AR152-22.tex

AR152-22.SGM

LaTeX2e(2001/05/10)

P1: GJC

CPG MOTIFS

Annu. Rev. Immunol. 2002.20:709-760. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

73.

74.

75.

76.

77.

78.

79.

80.

Stereo-enriched phosphorothioate oligodeoxynucleotides: synthesis, biophysical and biological properties. Bioorg. Med. Chem. 8(1):275–84 Yamamoto S, Yamamoto T, Kataoka T, Kuramoto E, Yano O, Tokunaga T. 1992. Unique palindromic sequences in synthetic oligonucleotides are required to induce IFN [correction of INF] and augment IFN-mediated [correction of INF] natural killer activity. J. Immunol. 148(12):4072– 76 Sun S, Zhang X, Tough DF, Sprent J. 1998. Type I interferon-mediated stimulation of T cells by CpG DNA. J. Exp. Med. 188(12):2335–42 Lipford GB, Bauer M, Blank C, Reiter R, Wagner H, Heeg K. 1997. CpG-containing synthetic oligonucleotides promote B and cytotoxic T cell responses to protein antigen: a new class of vaccine adjuvants. Eur. J. Immunol. 27(9):2340–44 Lipford GB, Bendigs S, Heeg K, Wagner H. 2000. Poly-guanosine motifs costimulate antigen-reactive CD8 T cells while bacterial CpG-DNA affect T-cell activation via antigen-presenting cell-derived cytokines. Immunology 101(1):46–52 Verthelyi D, Ishii KJ, Gursel M, Takeshita F, Klinman DM. 2001. Human peripheral blood cells differentially recognize and respond to two distinct CpG motifs. J. Immunol. 166(4):2372–77 Kadowaki N, Antonenko S, Liu YJ. 2001. Distinct CpG DNA and polyinosinicpolycytidylic acid double-stranded RNA, respectively, stimulate CD11c(−) type 2 dendritic cell precursors and CD11c(+) dendritic cells to produce type I IFN. J. Immunol. 166(4):2291–95 Sester DP, Naik S, Beasley SJ, Hume DA, Stacey KJ. 2000. Phosphorothioate backbone modification modulates macrophage activation by CpG DNA. J. Immunol. 165(8):4165–73 Vallin H, Perers A, Alm GV, Ronnblom L. 1999. Anti-double-stranded DNA antibodies and immunostimulatory plasmid

81.

82. 83.

84.

85.

86.

87.

88.

89.

747

DNA in combination mimic the endogenous IFN-α inducer in systemic lupus erythematosus. J. Immunol. 163(11):6306– 13 Krug A, Rothenfusser S, Hornung V, Jahrsdorfer B, Blackwell S, Ballas ZK, Endres S, Krieg AM, Hartmann G. 2001. Identification of CpG oligonucleotide sequences with high induction of IFNalpha/beta in plasmacytoid dendritic cells. Eur. J. Immunol. 31(7):2154–63 Krieg AM. 2001. Now I know my CpGs. Trends Microbiol. 9(6):249–52 Kimura Y, Sonehara K, Kuramoto E, Makino T, Yamamoto S, Yamamoto T, Kataoka T, Tokunaga T. 1994. Binding of oligoguanylate to scavenger receptors is required for oligonucleotides to augment NK cell activity and induce IFN. J. Biochem. (Tokyo) 116(5):991–94 Neujahr DC, Reich CF, Pisetsky DS. 1999. Immunostimulatory properties of genomic DNA from different bacterial species. Immunobiology 200(1):106–19 Mui B, Raney SG, Semple SC, Hope MJ. 2001. Immune stimulation by a CpGcontaining oligodeoxynucleotide is enhanced when encapsulated and delivered in lipid particles. J. Pharmacol. Exp. Ther. 298(3):1185–92 Stacey KJ, Sweet MJ, Hume DA. 1996. Macrophages ingest and are activated by bacterial DNA. J. Immunol. 157(5):2116– 22 Lipford GB, Sparwasser T, Bauer M, Zimmermann S, Koch ES, Heeg K, Wagner H. 1997. Immunostimulatory DNA: sequence-dependent production of potentially harmful or useful cytokines. Eur. J. Immunol. 27(12):3420–26 Chen Y, Lenert P, Weeranta R, McCluskie MJ, Wu T, Davis HL, Krieg AM. 2001. Identification of methylated CpG motifs as inhibitors of the immune stimulatory CpG motifs. Gene Ther. 8:1024–32 Lenert P, Stunz LL, Yi AK, Krieg AM, Ashman RF. 2001. CpG stimulation of primary mouse B cells is blocked by

1 Feb 2002

9:54

748

Annu. Rev. Immunol. 2002.20:709-760. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

90.

91.

92.

93.

94.

95.

96.

97.

AR

AR152-22.tex

AR152-22.SGM

LaTeX2e(2001/05/10)

P1: GJC

KRIEG inhibitory oligodeoxyribonucleotides at a site proximal to NF-kB activation. Antisense Nucleic Acid Drug Dev. 11(4):247– 56 Sato Y, Roman M, Tighe H, Lee D, Corr M, Nguyen MD, Silverman GJ, Lotz M, Carson DA, Raz E. 1996. Immunostimulatory DNA sequences necessary for effective intradermal gene immunization. Science 273(5273):352–54 Biswas S, Ashok MS, Reddy GS, Srinivasan VA, Rangarajan PN. 1999. Evaluation of the protective efficacy of rabies DNA vaccine in mice using an intracerebral challenge model. Curr. Sci. 76:1012– 16 Vinner L, Nielsen HV, Bryder K, Corbet S, Nielsen C, Fomsgaard A. 1999. Gene gun DNA vaccination with Revindependent synthetic HIV-1 gp160 envelope gene using mammalian codons. Vaccine 17(17):2166–75 Yew NS, Wang KX, Przybylska M, Bagley RG, Stedman M, Marshall J, Scheule RK, Cheng SH. 1999. Contribution of plasmid DNA to inflammation in the lung after administration of cationic lipid:pDNA complexes. Hum. Gene Ther. 10(2):223–34 Li S, Wu SP, Whitmore M, Loeffert EJ, Wang L, Watkins SC, Pitt BR, Huang L. 1999. Effect of immune response on gene transfer to the lung via systemic administration of cationic lipidic vectors. Am. J. Physiol. 276 (5, Pt. 1):L796–804 Yi AK, Klinman DM, Martin TL, Matson S, Krieg AM. 1996. Rapid immune activation by CpG motifs in bacterial DNA. Systemic induction of IL-6 transcription through an antioxidant-sensitive pathway. J. Immunol. 157(12):5394–402 Redford TW, Yi AK, Ward CT, Krieg AM. 1998. Cyclosporin A enhances IL12 production by CpG motifs in bacterial DNA and synthetic oligodeoxynucleotides. J. Immunol. 161(8):3930–35 Anitescu M, Chace JH, Tuetken R, Yi AK, Berg DJ, Krieg AM, Cowdery JS.

98.

99.

100.

101.

102.

103.

104.

1997. Interleukin-10 functions in vitro and in vivo to inhibit bacterial DNAinduced secretion of interleukin-12. J. Interferon Cytokine Res. 17(12):781–88 Cowdery JS, Chace JH, Yi AK, Krieg AM. 1996. Bacterial DNA induces NK cells to produce IFN-γ in vivo and increases the toxicity of lipopolysaccharides. J. Immunol. 156(12):4570–75 Yi AK, Chace JH, Cowdery JS, Krieg AM. 1996. IFN-γ promotes IL-6 and IgM secretion in response to CpG motifs in bacterial DNA and oligodeoxynucleotides. J. Immunol. 156(2):558–64 Davis HL, Weeratna R, Waldschmidt TJ, Tygrett L, Schorr J, Krieg AM, Weeratna R. 1998. CpG DNA is a potent enhancer of specific immunity in mice immunized with recombinant hepatitis B surface antigen. J. Immunol. 160(2):870–76 [Erratum J. Immunol. 1999, Mar. 1. 162 (5):3103] Kobayashi H, Horner AA, Takabayashi K, Nguyen MD, Huang E, Cinman N, Raz E. 1999. Immunostimulatory DNA pre-priming: a novel approach for prolonged Th1-biased immunity. Cell Immunol. 198(1):69–75 Decker T, Schneller F, Sparwasser T, Tretter T, Lipford GB, Wagner H, Peschel C. 2000. Immunostimulatory CpGoligonucleotides cause proliferation, cytokine production, and an immunogenic phenotype in chronic lymphocytic leukemia B cells. Blood 95(3):999–1006 Goeckeritz BE, Flora M, Witherspoon K, Vos Q, Lees A, Dennis GJ, Pisetsky DS, Klinman DM, Snapper CM, Mond JJ. 1999. Multivalent cross-linking of membrane Ig sensitizes murine B cells to a broader spectrum of CpG-containing oligodeoxynucleotide motifs, including their methylated counterparts, for stimulation of proliferation and Ig secretion. Int. Immunol. 11(10):1693–1700 Yi AK, Krieg AM. 1998. Rapid induction of mitogen-activated protein kinases by immune stimulatory CpG DNA. J. Immunol. 161(9):4493–97

1 Feb 2002

9:54

AR

AR152-22.tex

AR152-22.SGM

LaTeX2e(2001/05/10)

P1: GJC

Annu. Rev. Immunol. 2002.20:709-760. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

CPG MOTIFS 105. Yi AK, Hornbeck P, Lafrenz DE, Krieg AM. 1996. CpG DNA rescue of murine B lymphoma cells from anti-IgM-induced growth arrest and programmed cell death is associated with increased expression of c-myc and bcl-xL. J. Immunol. 157(11):4918–25 106. Yi AK, Krieg AM. 1998. CpG DNA rescue from anti-IgM-induced WEHI-231 B lymphoma apoptosis via modulation of IκBα and IκBβ and sustained activation of nuclear factor-κ B/c-Rel. J. Immunol. 160(3):1240–45 107. Han SS, Chung ST, Robertson DA, Chelvarajan RL, Bondada S. 1999. CpG oligodeoxynucleotides rescue BKS2 immature B cell lymphoma from anti-IgM-mediated growth inhibition by up-regulation of egr-1. Int. Immunol. 11(6):871–79 108. Yi AK, Peckham DW, Ashman RF, Krieg AM. 1999. CpG DNA rescues B cells from apoptosis by activating NFκB and preventing mitochondrial membrane potential disruption via a chloroquinesensitive pathway. Int. Immunol. 11(12): 2015–24 109. Wang Z, Karras JG, Colarusso TP, Foote LC, Rothstein TL. 1997. Unmethylated CpG motifs protect murine B lymphocytes against Fas-mediated apoptosis. Cell Immunol. 180(2):162–67 110. Krug A, Towarowski A, Britsch S, Rothenfusser S, Hornung V, Bals R, Giese T, Engelmann H, Endres S, Krieg AM, Hartmann G. 2001. Toll-like receptor expression reveals CpG DNA as a unique microbial stimulus for plasmacytoid dendritic cells which synergizes with CD40 ligand to induce high amounts of IL-12. Eur. J. Immunol. 31:3026–37 111. Kadowaki N, Ho S, Antonenko S, de Waal MR, Kastelein RA, Bazan F, Liu YJ. 2001. Subsets of human dendritic cell precursors express different toll-like receptors and respond to different microbial antigens. J. Exp. Med. 194(6):863–70 112. Bohle B, Jahn-Schmid B, Maurer D,

113.

114.

115.

116.

117.

118.

119.

749

Kraft D, Ebner C. 1999. Oligodeoxynucleotides containing CpG motifs induce IL-12, IL-18 and IFN-γ production in cells from allergic individuals and inhibit IgE synthesis in vitro. Eur. J. Immunol. 29(7): 2344–53 Schattenberg D, Schott M, Reindl G, Krueger T, Tschoepe D, Feldkamp J, Scherbaum WA, Seissler J. 2000. Response of human monocyte-derived dendritic cells to immunostimulatory DNA. Eur. J. Immunol. 30(10):2824–31 Bauer M, Redecke V, Ellwart JW, Scherer B, Kremer JP, Wagner H, Lipford GB. 2001. Bacterial CpG-DNA triggers activation and maturation of human CD11c(−), CD123(+) dendritic cells. J. Immunol. 166(8):5000–7 Jakob T, Walker PS, Krieg AM, Udey MC, Vogel JC. 1998. Activation of cutaneous dendritic cells by CpG-containing oligodeoxynucleotides: a role for dendritic cells in the augmentation of Th1 responses by immunostimulatory DNA. J. Immunol. 161(6):3042–49 Sparwasser T, Koch ES, Vabulas RM, Heeg K, Lipford GB, Ellwart JW, Wagner H. 1998. Bacterial DNA and immunostimulatory CpG oligonucleotides trigger maturation and activation of murine dendritic cells. Eur. J. Immunol. 28(6):2045–54 Ban E, Dupre L, Hermann E, Rohn W, Vendeville C, Quatannens B, RicciardiCastagnoli P, Capron A, Riveau G. 2000. CpG motifs induce Langerhans cell migration in vivo. Int. Immunol. 12(6):737– 45 Schulz O, Edwards AD, Schito M, Aliberti J, Manickasingham S, Sher A, Sousa C. 2000. CD40 triggering of heterodimeric IL-12 p70 production by dendritic cells in vivo requires a microbial priming signal. Immunity 13(4):453–62 Behboudi S, Chao D, Klenerman P, Austyn J. 2000. The effects of DNA containing CpG motif on dendritic cells. Immunology 99(3):361–66

1 Feb 2002

9:54

Annu. Rev. Immunol. 2002.20:709-760. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

750

AR

AR152-22.tex

AR152-22.SGM

LaTeX2e(2001/05/10)

P1: GJC

KRIEG

120. Hartmann G, Krieg AM. 1999. CpG DNA and LPS induce distinct patterns of activation in human monocytes. Gene Ther. 6(5):893–903 121. Pichyangkul S, Yongvanitchit K, Kumarb U, Krieg AM, Heppner DG, Walsh DS. 2001. Whole blood cultures to assess the immunostimulatory activities of CpG oligodeoxynucleotides. J. Immunol. Methods 247(1–2):83–94 122. Hartmann G, Krug A, Eigler A, Moeller J, Murphy J, Albrecht R, Endres S. 1996. Specific suppression of human tumor necrosis factor-α synthesis by antisense oligodeoxynucleotides. Antisense Nucleic Acid Drug Dev. 6(4):291–99 123. Schluesener HJ, Seid K, Deininger M, Schwab J. 2001. Transient in vivo activation of rat brain macrophages/microglial cells and astrocytes by immunostimulatory multiple CpG oligonucleotides. J. Neuroimmunol. 113(1):89–94 124. Deng GM, Liu ZQ, Tarkowski A. 2001. Intracisternally localized bacterial DNA containing CpG motifs induces meningitis. J. Immunol. 167(8):4616–26 125. Sweet MJ, Stacey KJ, Kakuda DK, Markovich D, Hume DA. 1998. IFN-γ primes macrophage responses to bacterial DNA. J. Interferon Cytokine Res. 18(4):263–71 126. Cowdery JS, Boerth NJ, Norian LA, Myung PS, Koretzky GA. 1999. Differential regulation of the IL-12 p40 promoter and of p40 secretion by CpG DNA and lipopolysaccharide. J. Immunol. 162(11):6770–75 127. Gao JJ, Zuvanich EG, Xue Q, Horn DL, Silverstein R, Morrison DC. 1999. Cutting edge: bacterial DNA and LPS act in synergy in inducing nitric oxide production in RAW 264.7 macrophages. J. Immunol. 163(8):4095–99 128. Yi AK, Yoon J-G, Hong S-C, Redford T, Krieg AM. 2001. Lipopolysaccharide and CpG DNA synergize for tumor necrosis factor-γ production through activation of NF-κB. Int. Immunol. 13(11):101–14

129. Gao JJ, Xue Q, Papasian CJ, Morrison DC. 2001. Bacterial DNA and lipopolysaccharide induce synergistic production of TNF-α through a post-transcriptional mechanism. J. Immunol. 166(11): 6855–60 130. Chen Y, Zhang J, Moore SA, Ballas ZK, Portanova JP, Krieg AM, Berg DJ. 2001. CpG DNA induces cyclooxygenase-2 expression and prostaglandin production. Int. Immunol. 13(8):1013–20 131. Chu RS, Askew D, Noss EH, Tobian A, Krieg AM, Harding CV. 1999. CpG oligodeoxynucleotides down-regulate macrophage class II MHC antigen processing. J. Immunol. 163(3):1188–94 132. Ramachandra L, Chu RS, Askew D, Noss EH, Canaday DH, Potter NS, Johnsen A, Krieg AM, Nedrud JG, Boom WH, Harding CV. 1999. Phagocytic antigen processing and effects of products on antigen processing and T-cell responses. Immunol. Rev. 168:217–39 133. Baek KH, Ha SJ, Sung YC. 2001. A novel function of phosphorothioate oligodeoxynucleotides as chemoattractants for primary macrophages. J. Immunol. 167(5):2847–54 134. Macfarlane DE, Manzel L. 1999. Immunostimulatory CpG-oligodeoxynucleotides induce a factor that inhibits macrophage adhesion. J. Lab Clin. Med. 134(5):501–9 135. Shimada S, Yano O, Tokunaga T. 1986. In vivo augmentation of natural killer cell activity with a deoxyribonucleic acid fraction of BCG. Jpn. J. Cancer Res. 77(8):808–16 136. Tokunaga T, Yano O, Kuramoto E, Kimura Y, Yamamoto T, Kataoka T, Yamamoto S. 1992. Synthetic oligonucleotides with particular base sequences from the cDNA encoding proteins of Mycobacterium bovis BCG induce interferons and activate natural killer cells. Microbiol. Immunol. 36(1):55–66 137. Kataoka T, Yamamoto S, Yamamoto T, Kuramoto E, Kimura Y, Yano O,

1 Feb 2002

9:54

AR

AR152-22.tex

AR152-22.SGM

LaTeX2e(2001/05/10)

P1: GJC

CPG MOTIFS

Annu. Rev. Immunol. 2002.20:709-760. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

138.

139.

140.

141.

142.

143.

144.

145.

Tokunaga T. 1992. Antitumor activity of synthetic oligonucleotides with sequences from cDNA encoding proteins of Mycobacterium bovis BCG. Jpn. J. Cancer Res. 83(3):244–47 Yamamoto T, Yamamoto S, Kataoka T, Tokunaga T. 1994. Ability of oligonucleotides with certain palindromes to induce interferon production and augment natural killer cell activity is associated with their base length. Antisense Res. Dev. 4(2):119–22 Yamamoto T, Yamamoto S, Kataoka T, Komuro K, Kohase M, Tokunaga T. 1994. Synthetic oligonucleotides with certain palindromes stimulate interferon production of human peripheral blood lymphocytes in vitro. Jpn. J. Cancer Res. 85(8):775–79 Wagner H. 1999. Bacterial CpG DNA activates immune cells to signal infectious danger. Adv. Immunol. 73:329–68 Bendigs S, Salzer U, Lipford GB, Wagner H, Heeg K. 1999. CpG-oligodeoxynucleotides co-stimulate primary T cells in the absence of antigen-presenting cells. Eur. J. Immunol. 29(4):1209–18 Lipford GB, Sparwasser T, Zimmermann S, Heeg K, Wagner H. 2000. CpG-DNAmediated transient lymphadenopathy is associated with a state of Th1 predisposition to antigen-driven responses. J. Immunol. 165(3):1228–35 Kranzer K, Bauer M, Lipford GB, Heeg K, Wagner H, Lang R. 2000. CpG-oligodeoxynucleotides enhance T-cell receptor-triggered interferon-γ production and up-regulation of CD69 via induction of antigen-presenting cell-derived interferon type I and interleukin-12. Immunology 99(2):170–78 Neufert C, Pai RK, Noss EH, Berger M, Boom WH, Harding CV. 2001. Mycobacterium tuberculosis 19-kDa Lipoprotein Promotes Neutrophil Activation. J. Immunol. 167(3):1542–49 Weighardt H, Feterowski C, Veit M, Rump M, Wagner H, Holzmann B. 2000.

146.

147.

148.

149.

150.

151.

152.

153.

751

Increased resistance against acute polymicrobial sepsis in mice challenged with immunostimulatory CpG oligodeoxynucleotides is related to an enhanced innate effector cell response. J. Immunol. 165(8):4537–43 Carlow DA, Teh SJ, Teh HS. 1998. Specific antiviral activity demonstrated by TGTP, a member of a new family of interferon-induced GTPases. J. Immunol. 161(5):2348–55 Takeshita A, Imai K, Hanazawa S. 1999. CpG motifs in Porphyromonas gingivalis DNA stimulate interleukin-6 expression in human gingival fibroblasts. Infect. Immun. 67(9):4340–45 Bauer S, Kirschning CJ, Hacker H, Redecke V, Hausmann S, Akira S, Wagner H, Lipford GB. 2001. Human TLR9 confers responsiveness to bacterial DNA via species-specific CpG motif recognition. Proc. Natl. Acad. Sci. USA 98(16):9237– 42 Klinman DM, Yi AK, Beaucage SL, Conover J, Krieg AM. 1996. CpG motifs present in bacteria DNA rapidly induce lymphocytes to secrete interleukin 6, interleukin 12, and interferon γ . Proc. Natl. Acad. Sci. USA 93(7):2879–83 Krieg AM, Love-Homan L, Yi AK, Harty JT. 1998. CpG DNA induces sustained IL12 expression in vivo and resistance to Listeria monocytogenes challenge. J. Immunol. 161(5):2428–34 Akbari O, Panjwani N, Garcia S, Tascon R, Lowrie D, Stockinger B. 1999. DNA vaccination: transfection and activation of dendritic cells as key events for immunity. J. Exp. Med. 189(1):169–78 Hacker H, Mischak H, Miethke T, Liptay S, Schmid R, Sparwasser T, Heeg K, Lipford GB, Wagner H. 1998. CpG-DNAspecific activation of antigen-presenting cells requires stress kinase activity and is preceded by non-specific endocytosis and endosomal maturation. EMBO J. 17(21):6230–40 Hughes JA, Avrutskaya AV, Juliano RL.

1 Feb 2002

9:54

752

Annu. Rev. Immunol. 2002.20:709-760. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

154.

155.

156.

157.

158.

159.

160.

161.

162.

AR

AR152-22.tex

AR152-22.SGM

LaTeX2e(2001/05/10)

P1: GJC

KRIEG 1994. Influence of base composition on membrane binding and cellular uptake of 10-mer phosphorothioate oligonucleotides in Chinese hamster ovary (CHRC5) cells. Antisense Res. Dev. 4(3): 211–15 Yao GQ, Corrias S, Cheng YC. 1996. Identification of two oligodeoxyribonucleotide binding proteins on plasma membranes of human cell lines. Biochem. Pharmacol. 51(4):431–36 Liang H, Reich CF, Pisetsky DS, Lipsky PE. 2000. The role of cell surface receptors in the activation of human B cells by phosphorothioate oligonucleotides. J. Immunol. 165(3):1438–45 Manzel L, Macfarlane DE. 1999. Lack of immune stimulation by immobilized CpG-oligodeoxynucleotide. Antisense Nucleic Acid Drug Dev. 9(5):459–64 Pisetsky DS, Reich CF III. 1999. The influence of lipofectin on the in vitro stimulation of murine spleen cells by bacterial DNA and plasmid DNA vectors. J. Interferon Cytokine Res. 19(10):1219–26 Gursel I, Gursel M, Ishii KJ, Klinman DM. 2001. Sterically stabilized cationic liposomes improve the uptake and immunostimulatory activity of CpG oligonucleotides. J. Immunol. 167(6):3324–28 Krieg AM, Gmelig-Meyling F, Gourley MF, Kisch WJ, Chrisey LA, Steinberg AD. 1991. Uptake of oligodeoxyribonucleotides by lymphoid cells is heterogeneous and inducible. Antisense Res. Dev. 1(2):161–71 Zhao Q, Song X, Waldschmidt T, Fisher E, Krieg AM. 1996. Oligonucleotide uptake in human hematopoietic cells is increased in leukemia and is related to cellular activation. Blood 88(5):1788–95 Kronenwett R, Steidl U, Kirsch M, Sczakiel G, Haas R. 1998. Oligodeoxyribonucleotide uptake in primary human hematopoietic cells is enhanced by cationic lipids and depends on the hematopoietic cell subset. Blood 91(3):852–62 Hartmann G, Krug A, Bidlingmaier M,

163.

164.

165.

166.

167.

168.

169.

170.

Hacker U, Eigler A, Albrecht R, Strasburger CJ, Endres S. 1998. Spontaneous and cationic lipid-mediated uptake of antisense oligonucleotides in human monocytes and lymphocytes. J. Pharmacol. Exp. Ther. 285(2):920–28 Britigan BE, Lewis TS, Waldschmidt M, McCormick ML, Krieg AM. 2001. Lactoferrin binds CpG-containing oligonucleotides and inhibits their immunostimulatory effects on human B cells. J. Immunol. 167(5):2921–28 Hanss B, Leal-Pinto E, Bruggeman LA, Copeland TD, Klotman PE. 1998. Identification and characterization of a cell membrane nucleic acid channel. Proc. Natl. Acad. Sci. USA 95(4):1921–26 Krieg AM. 1995. Uptake and localization of phosphodiester and chimeric oligodeoxynucleotides in normal and leukemic primary cells. In Delivery Strategies for Antisense Oligonucleotide Therapeutics, ed. S Akkhtar, pp. 177–90. Boca Raton, FL: CRC Yakubov LA, Deeva EA, Zarytova VF, Ivanova EM, Ryte AS, Yurchenko LV, Vlassov VV. 1989. Mechanism of oligonucleotide uptake by cells: involvement of specific receptors? Proc. Natl. Acad. Sci. USA 86(17):6454–58 Beltinger C, Saragovi HU, Smith RM, LeSauteur L, Shah N, DeDionisio L, Christensen L, Raible A, Jarett L, Gewirtz AM. 1995. Binding, uptake, and intracellular trafficking of phosphorothioatemodified oligodeoxynucleotides. J. Clin. Invest. 95(4):1814–23 Pichon C, Monsigny M, Roche AC. 1999. Intracellular localization of oligonucleotides: influence of fixative protocols. Antisense Nucleic Acid Drug Dev. 9(1):89–93 Clarenc JP, Lebleu B, Leonetti JP. 1993. Characterization of the nuclear binding sites of oligodeoxyribonucleotides and their analogs. J. Biol. Chem. 268(8):5600– 4 Chin DJ, Green GA, Zon G, Szoka FC Jr,

1 Feb 2002

9:54

AR

AR152-22.tex

AR152-22.SGM

LaTeX2e(2001/05/10)

P1: GJC

CPG MOTIFS

171.

Annu. Rev. Immunol. 2002.20:709-760. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

172.

173.

174.

175.

176.

177.

178.

Straubinger RM. 1990. Rapid nuclear accumulation of injected oligodeoxyribonucleotides. New Biol. 2(12):1091–100 Leonetti JP, Mechti N, Degols G, Gagnor C, Lebleu B. 1991. Intracellular distribution of microinjected antisense oligonucleotides. Proc. Natl. Acad. Sci. USA 88(7):2702–6 Morishita R, Gibbons GH, Ellison KE, Nakajima M, von der Leyen H, Zhang L, Kaneda Y, Ogihara T, Dzau VJ. 1994. Intimal hyperplasia after vascular injury is inhibited by antisense cdk 2 kinase oligonucleotides. J. Clin. Invest. 93(4):1458–64 Chiang MY, Chan H, Zounes MA, Freier SM, Lima WF, Bennett CF. 1991. Antisense oligonucleotides inhibit intercellular adhesion molecule 1 expression by two distinct mechanisms. J. Biol. Chem. 266(27):18,162–71 Nicklin PS, Craig SJ, Phillips JA. 1998. Pharmacokinetic properties of phosphorothioates in animals—absorption, distribution, metabolism and elimination. In Antisense Research and Application, ed. ST Crooke, 4:141–68. Berlin: Springer Sands H, Gorey-Feret LJ, Cocuzza AJ, Hobbs FW, Chidester D, Trainor GL. 1994. Biodistribution and metabolism of internally 3H-labeled oligonucleotides. I. Comparison of a phosphodiester and a phosphorothioate. Mol. Pharmacol. 45(5):932–43 Zhao Q, Zhou R, Temsamani J, Zhang Z, Roskey A, Agrawal S. 1998. Cellular distribution of phosphorothioate oligonucleotide following intravenous administration in mice. Antisense Nucleic Acid Drug Dev. 8(6):451–58 Yi AK, Tuetken R, Redford T, Waldschmidt M, Kirsch J, Krieg AM. 1998. CpG motifs in bacterial DNA activate leukocytes through the pH-dependent generation of reactive oxygen species. J. Immunol. 160(10):4755–61 Macfarlane DE, Manzel L. 1998. Antagonism of immunostimulatory CpG-

179.

180.

181.

182.

183.

184.

185.

186.

753

oligodeoxynucleotides by quinacrine, chloroquine, and structurally related compounds. J. Immunol. 160(3):1122– 31 Strekowski L, Zegrocka O, Henary M, Say M, Mokrosz MJ, Kotecka BM, Manzel L, Macfarlane DE. 1999. Structure-activity relationship analysis of substituted 4-quinolinamines, antagonists of immunostimulatory CpGoligodeoxynucleotides. Bioorg. Med. Chem. Lett. 9(13):1819–24 Hemmi H, Takeuchi O, Kawai T, Kaisho T, Sato S, Sanjo H, Matsumoto M, Hoshino K, Wagner H, Takeda K, Akira S. 2000. A toll-like receptor recognizes bacterial DNA. Nature 408(6813):740–45 Chuang TH, Ulevitch RJ. 2000. Cloning and characterization of a sub-family of human toll-like receptors: hTLR7, hTLR8 and hTLR9. Eur. Cytokine Netw. 11(3):372–78 Du X, Poltorak A, Wei Y, Beutler B. 2000. Three novel mammalian toll-like receptors: gene structure, expression, and evolution. Eur. Cytokine Netw. 11(3):362– 71 Takeshita F, Leifer C, Gursel I, Ishii KJ, Takeshita S, Gursel M, Klinman DM. 2001. Role of toll-like Receptor 9 in CpG DNA-induced activation of human cells. J. Immunol. 167:3555–58 Hacker H, Mischak H, Hacker G, Eser S, Prenzel N, Ullrich A, Wagner H. 1999. Cell type-specific activation of mitogenactivated protein kinases by CpG-DNA controls interleukin-12 release from antigen-presenting cells. EMBO J. 18(24): 6973–82 Dalpke AH, Opper S, Zimmermann S, Heeg K. 2001. Suppressors of cytokine signaling (SOCS)-1 and SOCS-3 are induced by CpG-DNA and modulate cytokine responses in APCs. J. Immunol. 166(12):7082–89 Sparwasser T, Miethke T, Lipford G, Erdmann A, Hacker H, Heeg K, Wagner H. 1997. Macrophages sense pathogens

1 Feb 2002

9:54

754

187.

Annu. Rev. Immunol. 2002.20:709-760. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

188.

189.

190.

191.

192.

193.

194.

195.

196.

AR

AR152-22.tex

AR152-22.SGM

LaTeX2e(2001/05/10)

P1: GJC

KRIEG via DNA motifs: induction of tumor necrosis factor-α-mediated shock. Eur. J. Immunol. 27(7):1671–79 Chu W, Gong X, Li Z, Takabayashi K, Ouyang H, Chen Y, Lois A, Chen DJ, Li GC, Karin M, Raz E. 2000. DNAPKcs is required for activation of innate immunity by immunostimulatory DNA. Cell 103(6):909–18 Gupta RK, Siber GR. 1995. Adjuvants for human vaccines—current status, problems and future prospects. Vaccine 13(14):1263–76 Bireland ML, Monroe JG. 1997. Biochemistry of antigen receptor signaling in mature and developing B lymphocytes. Crit. Rev. Immunol. 17(3–4):353–85 DeFranco AL. 1997. The complexity of signaling pathways activated by the BCR. Curr. Opin. Immunol. 9(3):296–308 Sweet MJ, Stacey KJ, Ross IL, Ostrowski MC, Hume DA. 1998. Involvement of Ets, rel and Sp1-like proteins in lipopolysaccharide-mediated activation of the HIV-1 LTR in macrophages. J. Inflamm. 48(2):67–83 Takeshita F, Ishii KJ, Ueda A, Ishigatsubo Y, Klinman DM. 2000. Positive and negative regulatory elements contribute to CpG oligonucleotide-mediated regulation of human IL-6 gene expression. Eur. J. Immunol. 30(1):108–16 Takeshita F, Klinman DM. 2000. CpG ODN-mediated regulation of IL-12 p40 transcription. Eur. J. Immunol. 30(7): 1967–76 Chace JH, Hooker NA, Mildenstein KL, Krieg AM, Cowdery JS. 1997. Bacterial DNA-induced NK cell IFN-γ production is dependent on macrophage secretion of IL-12. Clin. Immunol. Immunopathol. 84(2):185–93 Schwartz DA, Quinn TJ, Thorne PS, Sayeed S, Yi AK, Krieg AM. 1997. CpG motifs in bacterial DNA cause inflammation in the lower respiratory tract. J. Clin. Invest. 100(1):68–73 Roman M, Martin-Orozco E, Goodman

197.

198.

199.

200. 201.

202.

203.

204.

JS, Nguyen MD, Sato Y, Ronaghy A, Kornbluth RS, Richman DD, Carson DA, Raz E. 1997. Immunostimulatory DNA sequences function as T helper-1-promoting adjuvants. Nat. Med. 3(8):849–54 Zhao Q, Temsamani J, Zhou RZ, Agrawal S. 1997. Pattern and kinetics of cytokine production following administration of phosphorothioate oligonucleotides in mice. Antisense Nucleic Acid Drug Dev. 7(5):495–502 Waag DM, Kende M, Damrow TA, Wood OL, Williams JC. 1990. Injection of inactivated phase I Coxiella burnetii increases non-specific resistance to infection and stimulates lymphokine production in mice. Ann. N. Y. Acad. Sci. 590:203–14 Vogels MT, van der Meer JW. 1992. Use of immune modulators in nonspecific therapy of bacterial infections. Antimicrob. Agents Chemother. 36(1):1–5 Hadden JW. 1993. Immunostimulants. Trends Pharmacol. Sci. 14(5):169–74 Oxenius A, Martinic MM, Hengartner H, Klenerman P. 1999. CpG-containing oligonucleotides are efficient adjuvants for induction of protective antiviral immune responses with T-cell peptide vaccines. J. Virol. 73(5):4120–26 Elkins KL, Rhinehart-Jones TR, Stibitz S, Conover JS, Klinman DM. 1999. Bacterial DNA containing CpG motifs stimulates lymphocyte-dependent protection of mice against lethal infection with intracellular bacteria. J. Immunol. 162(4):2291– 98 Klinman DM, Kamstrup S, Verthelyi D, Gursel I, Ishii KJ, Takeshita F, Gursel M. 2000. Activation of the innate immune system by CpG oligodeoxynucleotides: immunoprotective activity and safety. Springer Semin. Immunopathol. 22(1–2):173–83 Klinman DM, Conover J, Coban C. 1999. Repeated administration of synthetic oligodeoxynucleotides expressing CpG motifs provides long-term protection

1 Feb 2002

9:54

AR

AR152-22.tex

AR152-22.SGM

LaTeX2e(2001/05/10)

P1: GJC

CPG MOTIFS

205.

Annu. Rev. Immunol. 2002.20:709-760. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

206.

207.

208.

209.

210.

211.

212.

213.

against bacterial infection. Infect. Immun. 67(11):5658–63 Zimmermann S, Egeter O, Hausmann S, Lipford GB, Rocken M, Wagner H, Heeg K. 1998. CpG oligodeoxynucleotides trigger protective and curative Th1 responses in lethal murine leishmaniasis. J. Immunol. 160(8):3627–30 Walker PS, Scharton-Kersten T, Krieg AM, Love-Homan L, Rowton ED, Udey MC, Vogel JC. 1999. Immunostimulatory oligodeoxynucleotides promote protective immunity and provide systemic therapy for leishmaniasis via IL-12-and IFNγ -dependent mechanisms. Proc. Natl. Acad. Sci. USA 96(12):6970–75 Stacey KJ, Blackwell JM. 1999. Immunostimulatory DNA as an adjuvant in vaccination against Leishmania major. Infect. Immun. 67(8):3719–26 Gramzinski RA, Sedegah M, Doolan D, Krieg AM, Davis HL, Hoffman SL. 2000. IL12- and IFN-γ -dependent protection of mice against malaria conferred by CpG DNA. Infect. Immun. 69:1643–49 Klinman DM, Verthelyi D, Takeshita F, Ishii KJ. 1999. Immune recognition of foreign DNA: a cure for bioterrorism? Immunity 11(2):123–29 Hatzfeld J, Li ML, Brown EL, Sookdeo H, Levesque JP, O’Toole T, Gurney C, Clark SC, Hatzfeld A. 1991. Release of early human hematopoietic progenitors from quiescence by antisense transforming growth factor β 1 or Rb oligonucleotides. J. Exp. Med. 174(4):925–29 Sparwasser T, Hultner L, Koch ES, Luz A, Lipford GB, Wagner H. 1999. Immunostimulatory CpG-oligodeoxynucleotides cause extramedullary murine hemopoiesis. J. Immunol. 162(4):2368– 74 Krieg AM, Davis HL. 2001. Enhancing vaccines with immune stimulatory CpG DNA. Curr. Opin. Mol. Ther. 3(1):15– 24 Shankaran V, Ikeda H, Bruce AT, White JM, Swanson PE, Old LJ, Schreiber RD.

214.

215.

216.

217.

218.

219.

220.

221.

755

2001. IFNgamma and lymphocytes prevent primary tumour development and shape tumour immunogenicity. Nature 410(6832):1107–11 Wickstrom E. 1997. Antisense c-myc inhibition of lymphoma growth. Antisense Nucleic Acid Drug Dev. 7(3):225–28 Smith JB, Wickstrom E. 1998. Antisense c-myc and immunostimulatory oligonucleotide inhibition of tumorigenesis in a murine B-cell lymphoma transplant model. J. Natl. Cancer Inst. 90(15):1146– 54 Dow SW, Fradkin LG, Liggitt DH, Willson AP, Heath TD, Potter TA. 1999. Lipid-DNA complexes induce potent activation of innate immune responses and antitumor activity when administered intravenously. J. Immunol. 163(3):1552– 61 Lanuti M, Rudginsky S, Force SD, Lambright ES, Siders WM, Chang MY, Amin KM, Kaiser LR, Scheule RK, Albelda SM. 2000. Cationic lipid:bacterial DNA complexes elicit adaptive cellular immunity in murine intraperitoneal tumor models. Cancer Res. 60(11):2955–63 Hafner M, Zawatzky R, Hirtreiter C, Buurman WA, Echtenacher B, Hehlgans T, Mannel DN. 2001. Antimetastatic effect of CpG DNA mediated by type I IFN. Cancer Res. 61(14):5523–28 Ballas Z, Krieg AM, Warren TL, Rasmussen WL, Davis HL, Waldschmidt M, Weiner GJ. 2001. Divergent therapeutic and immunologic effects of oligodeoxynucleotides with distinct CpG motifs. J. Immunol. 167:4878–86 Carpentier AF, Chen L, Maltonti F, Delattre JY. 1999. Oligodeoxynucleotides containing CpG motifs can induce rejection of a neuroblastoma in mice. Cancer Res. 59(21):5429–32 Carpentier AF, Xie J, Mokhtari K, Delattre JY. 2000. Successful treatment of intracranial gliomas in rat by oligodeoxynucleotides containing CpG motifs. Clin. Cancer Res. 6(6):2469–73

1 Feb 2002

9:54

Annu. Rev. Immunol. 2002.20:709-760. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

756

AR

AR152-22.tex

AR152-22.SGM

LaTeX2e(2001/05/10)

P1: GJC

KRIEG

222. Blazar BR, Krieg AM, Taylor PA. 2001. Synthetic unmethylated cytosinephosphate-guanosine oligodeoxynucleotides are potent stimulators of antileukemia responses in naive and bone marrow transplant recipients. Blood 98(4):1217–25 223. Sparwasser T, Vabulas RM, Villmow B, Lipford GB, Wagner H. 2000. Bacterial CpG-DNA activates dendritic cells in vivo: T helper cell-independent cytotoxic T cell responses to soluble proteins. Eur. J. Immunol. 30(12):3591–97 224. Vabulas RM, Pircher H, Lipford GB, Hacker H, Wagner H. 2000. CpG-DNA activates in vivo T cell epitope presenting dendritic cells to trigger protective antiviral cytotoxic T cell responses. J. Immunol. 164(5):2372–78 225. Chu RS, Targoni OS, Krieg AM, Lehmann PV, Harding CV. 1997. CpG oligodeoxynucleotides act as adjuvants that switch on T helper 1 (Th1) immunity. J. Exp. Med. 186(10):1623–31 226. Wild J, Grusby MJ, Schirmbeck R, Reimann J. 1999. Priming MHC-I-restricted cytotoxic T lymphocyte responses to exogenous hepatitis B surface antigen is CD4+ T cell dependent. J. Immunol. 163(4):1880–87 227. Tascon RE, Ragno S, Lowrie DB, Colston MJ. 2000. Immunostimulatory bacterial DNA sequences activate dendritic cells and promote priming and differentiation of CD8+ T cells. Immunology 99(1):1–7 228. Chan K, Lee DJ, Schubert A, Tang CM, Crain B, Schoenberger SP, Corr M. 2001. The roles of MHC class II, CD40, and B7 costimulation in CTL induction by plasmid DNA. J. Immunol. 166(5):3061– 66 229. Kim SK, Ragupathi G, Musselli C, Choi SJ, Park YS, Livingston PO. 1999. Comparison of the effect of different immunological adjuvants on the antibody and T-cell response to immunization with MUC1-KLH and GD3-KLH conjugate

230.

231.

232.

233.

234.

235.

236.

cancer vaccines. Vaccine 18(7–8):597– 603 Kim SK, Ragupathi G, Cappello S, Kagan E, Livingston PO. 2000. Effect of immunological adjuvant combinations on the antibody and T-cell response to vaccination with MUC1-KLH and GD3-KLH conjugates. Vaccine 19(4-5):530–37 Weiner GJ, Liu HM, Wooldridge JE, Dahle CE, Krieg AM. 1997. Immunostimulatory oligodeoxynucleotides containing the CpG motif are effective as immune adjuvants in tumor antigen immunization. Proc. Natl. Acad. Sci. USA 94(20):10,833–37 Liu HM, Newbrough SE, Bhatia SK, Dahle CE, Krieg AM, Weiner GJ. 1998. Immunostimulatory CpG oligodeoxynucleotides enhance the immune response to vaccine strategies involving granulocytemacrophage colony-stimulating factor. Blood 92(10):3730–36 Davila E, Celis E. 2000. Repeated administration of cytosine-phosphorothiolated guanine-containing oligonucleotides together with peptide/protein immunization results in enhanced CTL responses with anti-tumor activity. J. Immunol. 165(1):539–47 Egeter O, Mocikat R, Ghoreschi K, Dieckmann A, Rocken M. 2000. Eradication of disseminated lymphomas with CpG-DNA activated T helper type 1 cells from nontransgenic mice. Cancer Res. 60(6):1515–20 Brunner C, Seiderer J, Schlamp A, Bidlingmaier M, Eigler A, Haimerl W, Lehr HA, Krieg AM, Hartmann G, Endres S. 2000. Enhanced dendritic cell maturation by TNF-α or cytidine-phosphateguanosine DNA drives T cell activation in vitro and therapeutic anti-tumor immune responses in vivo. J. Immunol. 165(11):6278–86 Wooldridge JE, Ballas Z, Krieg AM, Weiner GJ. 1997. Immunostimulatory oligodeoxynucleotides containing CpG motifs enhance the efficacy of monoclonal

1 Feb 2002

9:54

AR

AR152-22.tex

AR152-22.SGM

LaTeX2e(2001/05/10)

P1: GJC

CPG MOTIFS

237.

Annu. Rev. Immunol. 2002.20:709-760. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

238.

239. 240.

241.

242.

243.

244.

245.

antibody therapy of lymphoma. Blood 89(8):2994–98 Warren TL, Dahle CE, Weiner GJ. 2000. CpG oligodeoxynucleotides enhance monoclonal antibody therapy of a murine lymphoma. Clin. Lymphoma 1:57– 61 Jahrsdorfer B, Hartmann G, Racila E, Jackson W, Muhlenhoff L, Meinhardt G, Endres S, Link BK, Krieg AM, Weiner GJ. 2001. CpG DNA increases primary malignant B cell expression of costimulatory molecules and target antigens. J. Leukocyte. Biol. 69(1):81–88 Deleted in proof Decker T, Schneller F, Kronschnabl M, Dechow T, Lipford GB, Wagner H, Peschel C. 2000. Immunostimulatory CpG-oligonucleotides induce functional high-affinity IL-2 receptors on B-CLL cells: Costimulation with IL-2 results in a highly immunogenic phenotype. Exp. Hematol. 28(5):558–68 Kline JN, Krieg AM, Waldschmidt TJ, Ballas ZK, Jain V, Businga TR. 1999. CpG oligodeoxynucleotides do not require TH1 cytokines to prevent eosinophilic airway inflammation in a murine model of asthma. J. Allergy Clin. Immunol. 104(6):1258–64 Erb KJ. 1999. Atopic disorders: a default pathway in the absence of infection? Immunol. Today 20(7):317–22 Sudo N, Aiba Y, Takaki A, Tanaka K, Yu XN, Oyama N, Koga Y, Kubo C. 2000. Dietary nucleic acids promote a shift in Th1/Th2 balance toward Th1-dominant immunity. Clin. Exp. Allergy 30(7):979– 87 Kuchan M, Winship T, Masor M. 1998. Nucleotides in infant nutrition: effects on immune function. In Pediatric Nutrition, ed. R Reifen, A. Lerner, D Branski, HSA Heymans, pp. 80–94. Basel: Karger Kline JN, Waldschmidt TJ, Businga TR, Lemish JE, Weinstock JV, Thorne PS, Krieg AM. 1998. Modulation of airway inflammation by CpG oligodeoxynu-

246.

247.

248.

249.

250.

251.

252.

757

cleotides in a murine model of asthma. J. Immunol. 160(6):2555–59 Lee DJ, Tighe H, Corr M, Roman M, Carson DA, Spiegelberg HL, Raz E. 1997. Inhibition of IgE antibody formation by plasmid DNA immunization is mediated by both CD4+ and CD8+ T cells. Int. Arch. Allergy Immunol. 113(1– 3):227–30 Chiaramonte MG, Hesse M, Cheever AW, Wynn TA. 2000. CpG oligonucleotides can prophylactically immunize against Th2-mediated schistosome egg-induced pathology by an IL-12independent mechanism. J. Immunol. 164(2):973–85 Fujieda S, Iho S, Kimura Y, Yamamoto H, Igawa H, Saito H. 2000. Synthetic oligodeoxynucleotides inhibit IgE induction in human lymphocytes. Am. J. Respir. Crit. Care Med. 162(1):232–39 Broide D, Schwarze J, Tighe H, Gifford T, Nguyen MD, Malek S, Van Uden J, Martin-Orozco E, Gelfand EW, Raz E. 1998. Immunostimulatory DNA sequences inhibit IL-5, eosinophilic inflammation, and airway hyperresponsiveness in mice. J. Immunol. 161(12):7054–62 Sur S, Wild JS, Choudhury BK, Sur N, Alam R, Klinman DM. 1999. Longterm prevention of allergic lung inflammation in a mouse model of asthma by CpG oligodeoxynucleotides. J. Immunol. 162(10):6284–93 Kohama Y, Akizuki O, Hagihara K, Yamada E, Yamamoto H. 1999. Immunostimulatory oligodeoxynucleotide induces TH1 immune response and inhibition of IgE antibody production to cedar pollen allergens in mice. J. Allergy Clin. Immunol. 104(6):1231–38 Hartl A, Kiesslich J, Weiss R, Bernhaupt A, Mostbock S, Scheiblhofer S, Ebner C, Ferreira F, Thalhamer J. 1999. Immune responses after immunization with plasmid DNA encoding Bet v 1, the major allergen of birch pollen. J. Allergy Clin. Immunol. 103(1 Pt. 1):107–13

1 Feb 2002

9:54

Annu. Rev. Immunol. 2002.20:709-760. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

758

AR

AR152-22.tex

AR152-22.SGM

LaTeX2e(2001/05/10)

P1: GJC

KRIEG

253. Jahn-Schmid B, Wiedermann U, Bohle B, Repa A, Kraft D, Ebner C. 1999. Oligodeoxynucleotides containing CpG motifs modulate the allergic TH2 response of BALB/c mice to Bet v 1, the major birch pollen allergen. J. Allergy Clin. Immunol. 104(5):1015–23 254. Serebrisky D, Teper AA, Huang CK, Lee SY, Zhang TF, Schofield BH, Kattan M, Sampson HA, Li XM. 2000. CpG oligodeoxynucleotides can reverse Th2-associated allergic airway responses and alter the B7.1/B7.2 expression in a murine model of asthma. J. Immunol. 165(10):5906–12 255. Isenberg DA, Ehrenstein MR, Longhurst C, Kalsi JK. 1994. The origin, sequence, structure, and consequences of developing anti-DNA antibodies. A human perspective. Arthritis Rheum. 37(2):169– 80 256. Pisetsky DS. 1996. The immunologic properties of DNA. J. Immunol. 156(2):421–23 257. Erikson J, Radic MZ, Camper SA, Hardy RR, Carmack C, Weigert M. 1991. Expression of anti-DNA immunoglobulin transgenes in non-autoimmune mice. Nature 349(6307):331–34 258. Chen C, Nagy Z, Radic MZ, Hardy RR, Huszar D, Camper SA, Weigert M. 1995. The site and stage of anti-DNA B-cell deletion. Nature 373(6511):252– 55 259. Chen C, Radic MZ, Erikson J, Camper SA, Litwin S, Hardy RR, Weigert M. 1994. Deletion and editing of B cells that express antibodies to DNA. J. Immunol. 152(4):1970–82 260. Roark JH, Kuntz CL, Nguyen KA, Caton AJ, Erikson J. 1995. Breakdown of B cell tolerance in a mouse model of systemic lupus erythematosus. J. Exp. Med. 181(3):1157–67 261. Ray SK, Putterman C, Diamond B. 1996. Pathogenic autoantibodies are routinely generated during the response to foreign antigen: a paradigm for autoim-

mune disease. Proc. Natl. Acad. Sci. USA 93(5):2019–24 262. Datta SK, Kaliyaperumal A. 1997. Nucleosome-driven autoimmune response in lupus. Pathogenic T helper cell epitopes and costimulatory signals. Ann. N. Y. Acad. Sci. 815:155–170 263. Krishnan MR, Marion TN. 1993. Structural similarity of antibody variable regions from immune and autoimmune anti-DNA antibodies. J. Immunol. 150 (11):4948–57 264. Desai DD, Krishnan MR, Swindle JT, Marion TN. 1993. Antigen-specific induction of antibodies against native mammalian DNA in nonautoimmune mice. J. Immunol. 151(3):1614–26 265. Moens U, Seternes OM, Hey AW, Silsand Y, Traavik T, Johansen B, Rekvig OP. 1995. In vivo expression of a single viral DNA-binding protein generates systemic lupus erythematosus–related autoimmunity to double-stranded DNA and histones. Proc. Natl. Acad. Sci. USA 92(26):12,393–97 266. Rekvig OP, Moens U, Sundsfjord A, Bredholt G, Osei A, Haaheim H, Traavik T, Arnesen E, Haga HJ. 1997. Experimental expression in mice and spontaneous expression in human SLE of polyomavirus T-antigen. A molecular basis for induction of antibodies to DNA and eukaryotic transcription factors. J. Clin. Invest. 99(8):2045–54 267. Gilkeson GS, Conover J, Halpern M, Pisetsky DS, Feagin A, Klinman DM. 1998. Effects of bacterial DNA on cytokine production by (NZB/NZW) F1 mice. J. Immunol. 161(8):3890– 95 268. Gilkeson GS, Pippen AM, Pisetsky DS. 1995. Induction of cross-reactive antidsDNA antibodies in preautoimmune NZB/NZW mice by immunization with bacterial DNA. J. Clin. Invest. 95(3):1398–1402 269. Gilkeson GS, Ruiz P, Pippen AM, Alexander AL, Lefkowith JB, Pisetsky

1 Feb 2002

9:54

AR

AR152-22.tex

AR152-22.SGM

LaTeX2e(2001/05/10)

P1: GJC

CPG MOTIFS

270.

Annu. Rev. Immunol. 2002.20:709-760. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

271.

272.

273.

274.

274a. 275.

276.

277.

DS. 1996. Modulation of renal disease in autoimmune NZB/NZW mice by immunization with bacterial DNA. J. Exp. Med. 183(4):1389–97 Mor G, Singla M, Steinberg AD, Hoffman SL, Okuda K, Klinman DM. 1997. Do DNA vaccines induce autoimmune disease? Hum. Gene Ther. 8(3):293–300 Anderson AC, Nicholson LB, Legge KL, Turchin V, Zaghouani H, Kuchroo VK. 2000. High frequency of autoreactive myelin proteolipid protein-specific T cells in the periphery of naive mice: mechanisms of selection of the self-reactive repertoire. J. Exp. Med. 191(5):761–70 Klein L, Klugmann M, Nave KA, Tuohy VK, Kyewski B. 2000. Shaping of the autoreactive T-cell repertoire by a splice variant of self protein expressed in thymic epithelial cells. Nat. Med. 6(1):56–61 Segal BM, Klinman DM, Shevach EM. 1997. Microbial products induce autoimmune disease by an IL-12-dependent pathway. J. Immunol. 158(11):5087–90 Segal BM, Chang JT, Shevach EM. 2000. CpG oligonucleotides are potent adjuvants for the activation of autoreactive encephalitogenic T cells in vivo. J. Immunol. 164(11):5683–88 Vishnupad N, Krieg AM, Cort L, Blankenhorn EP. 2002. Submitted Tsunoda I, Tolley ND, Theil DJ, Whitton JL, Kobayashi H, Fujinami RS. 1999. Exacerbation of viral and autoimmune animal models for multiple sclerosis by bacterial DNA. Brain Pathol. 9(3):481–93 Ruiz PJ, Garren H, Ruiz IU, Hirschberg DL, Nguyen LV, Karpuj MV, Cooper MT, Mitchell DJ, Fathman CG, Steinman L. 1999. Suppressive immunization with DNA encoding a self-peptide prevents autoimmune disease: modulation of T cell costimulation. J. Immunol. 162(6):3336–41 Lobell A, Weissert R, Storch MK,

278.

279.

280.

281.

282.

283.

284.

759

Svanholm C, de Graaf KL, Lassmann H, Andersson R, Olsson T, Wigzell H. 1998. Vaccination with DNA encoding an immunodominant myelin basic protein peptide targeted to Fc of immunoglobulin G suppresses experimental autoimmune encephalomyelitis. J. Exp. Med. 187(9):1543–48 Lobell A, Weissert R, Eltayeb S, Svanholm C, Olsson T, Wigzell H. 1999. Presence of CpG DNA and the local cytokine milieu determine the efficacy of suppressive DNA vaccination in experimental autoimmune encephalomyelitis. J. Immunol. 163(9):4754–62 Boccaccio GL, Mor F, Steinman L. 1999. Non-coding plasmid DNA induces IFN-γ in vivo and suppresses autoimmune encephalomyelitis. Int. Immunol. 11(2):289–96 Wildbaum G, Westermann J, Maor G, Karin N. 2000. A targeted DNA vaccine encoding fas ligand defines its dual role in the regulation of experimental autoimmune encephalomyelitis. J. Clin. Invest. 106(5):671–79 Youssef S, Wildbaum G, Maor G, Lanir N, Gour-Lavie A, Grabie N, Karin N. 1998. Long-lasting protective immunity to experimental autoimmune encephalomyelitis following vaccination with naked DNA encoding C-C chemokines. J. Immunol. 161(8):3870– 79 Wildbaum G, Youssef S, Karin N. 2000. A targeted DNA vaccine augments the natural immune response to self TNF-α and suppresses ongoing adjuvant arthritis. J. Immunol. 165(10):5860– 66 Bachmaier K, Neu N, de la Maza LM, Pal S, Hessel A, Penninger JM. 1999. Chlamydia infections and heart disease linked through antigenic mimicry. Science 283(5406):1335–39 Miyata M, Kobayashi H, Sasajima T, Sato Y, Kasukawa R. 2000. Unmethylated oligo-DNA containing CpG motifs

1 Feb 2002

9:54

760

285.

Annu. Rev. Immunol. 2002.20:709-760. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

286.

287.

287a.

288.

289.

290.

AR

AR152-22.tex

AR152-22.SGM

LaTeX2e(2001/05/10)

P1: GJC

KRIEG aggravates collagen-induced arthritis in. Arthritis Rheum. 43(11):2578–82 Kouskoff V, Lacaud G, Nemazee D. 2000. T cell-independent rescue of B lymphocytes from peripheral immune tolerance. Science 287(5462):2501–3 Mancini M, Hadchouel M, Davis HL, Whalen RG, Tiollais P, Michel ML. 1996. DNA-mediated immunization in a transgenic mouse model of the hepatitis B surface antigen chronic carrier state. Proc. Natl. Acad. Sci. USA 93(22):12,496–501 Mancini M, Hadchouel M, Tiollais P, Michel ML. 1998. Regulation of hepatitis B virus mRNA expression in a hepatitis B surface antigen transgenic mouse model by IFN-γ -secreting T cells after DNA-based immunization. J. Immunol. 161(10):5564–70 Malanchere-Bres E, Payette PJ, Mancini M, Tiollais P, Davis HL, Michel ML. 2001. CpG oligodeoxynucleotides with hepatitis B surface antigen (HBsAg) for vaccination in HBsAgtransgenic mice. J. Virol. 75(14):6482– 91 Deng GM, Nilsson IM, Verdrengh M, Collins LV, Tarkowski A. 1999. Intraarticularly localized bacterial DNA containing CpG motifs induces arthritis. Nat. Med. 5(6):702–5 Deng GM, Tarkowski A. 2000. The features of arthritis induced by CpG motifs in bacterial DNA. Arthritis Rheum. 43(2):356–64 Deng GM, Verdrengh M, Liu ZQ, Tarkowski A. 2000. The major role of macrophages and their product tumor necrosis factor α in the induction of arthritis triggered by bacterial DNA containing CpG motifs. Arthritis Rheum. 43(10):2283–89

291. Gerard HC, Wang Z, Wang GF, El Gabalawy H, Goldbach-Mansky R, Li Y, Majeed W, Zhang H, Ngai N, Hudson AP, Schumacher HR. 2001. Chromosomal DNA from a variety of bacterial species is present in synovial tissue from patients with various forms of arthritis. Arthritis Rheum. 44(7):1689– 97 292. Sparwasser T, Miethke T, Lipford G, Borschert K, Hacker H, Heeg K, Wagner H. 1997. Bacterial DNA causes septic shock. Nature 386(6623):336– 37 293. Schwartz DA, Wohlford-Lenane CL, Quinn TJ, Krieg AM. 1999. Bacterial DNA or oligonucleotides containing unmethylated CpG motifs can minimize lipopolysaccharide-induced inflammation in the lower respiratory tract through an IL-12-dependent pathway. J. Immunol. 163(1):224–31 294. Fredriksen K, Skogsholm A, Flaegstad T, Traavik T, Rekvig OP. 1993. Antibodies to dsDNA are produced during primary BK virus infection in man, indicating that anti-dsDNA antibodies may be related to virus replication in vivo. Scand. J. Immunol. 38(4):401–6 295. Krieg AM. 1999. Direct immunologic activities of CpG DNA and implications for gene therapy. J. Gene Med. 1(1):56– 63 296. Krieg AM. 2000. Minding the Cs and Gs. Mol. Ther. 1(3):209–10 297. Glover JM, Leeds JM, Mant TG, Amin D, Kisner DL, Zuckerman JE, Geary RS, Levin AA, Shanahan WR Jr. 1997. Phase I safety and pharmacokinetic profile of an intercellular adhesion molecule-1 antisense oligodeoxynucleotide (ISIS 2302). J. Pharmacol. Exp. Ther. 282(3):1173–80

P1: FRK

February 1, 2002

12:47

Annual Reviews

AR152-FM

Annual Review of Immunology Volume 20, 2002

Annu. Rev. Immunol. 2002.20:709-760. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

CONTENTS FRONTISPIECE, Charles A. Janeway, Jr. A TRIP THROUGH MY LIFE WITH AN IMMUNOLOGICAL THEME, Charles A. Janeway, Jr.

xiv 1

THE B7 FAMILY OF LIGANDS AND ITS RECEPTORS: NEW PATHWAYS FOR COSTIMULATION AND INHIBITION OF IMMUNE RESPONSES, Beatriz M. Carreno and Mary Collins

29

MAP KINASES IN THE IMMUNE RESPONSE, Chen Dong, Roger J. Davis, and Richard A. Flavell

PROSPECTS FOR VACCINE PROTECTION AGAINST HIV-1 INFECTION AND AIDS, Norman L. Letvin, Dan H. Barouch, and David C. Montefiori T CELL RESPONSE IN EXPERIMENTAL AUTOIMMUNE ENCEPHALOMYELITIS (EAE): ROLE OF SELF AND CROSS-REACTIVE ANTIGENS IN SHAPING, TUNING, AND REGULATING THE AUTOPATHOGENIC T CELL REPERTOIRE, Vijay K. Kuchroo, Ana C. Anderson, Hanspeter Waldner, Markus Munder, Estelle Bettelli, and Lindsay B. Nicholson

55 73

101

NEUROENDOCRINE REGULATION OF IMMUNITY, Jeanette I. Webster, Leonardo Tonelli, and Esther M. Sternberg

125

MOLECULAR MECHANISM OF CLASS SWITCH RECOMBINATION: LINKAGE WITH SOMATIC HYPERMUTATION, Tasuku Honjo, Kazuo Kinoshita, and Masamichi Muramatsu

165

INNATE IMMUNE RECOGNITION, Charles A. Janeway, Jr. and Ruslan Medzhitov

KIR: DIVERSE, RAPIDLY EVOLVING RECEPTORS OF INNATE AND ADAPTIVE IMMUNITY, Carlos Vilches and Peter Parham ORIGINS AND FUNCTIONS OF B-1 CELLS WITH NOTES ON THE ROLE OF CD5, Robert Berland and Henry H. Wortis E PROTEIN FUNCTION IN LYMPHOCYTE DEVELOPMENT, Melanie W. Quong, William J. Romanow, and Cornelis Murre

197 217 253 301

LYMPHOCYTE-MEDIATED CYTOTOXICITY, John H. Russell and Timothy J. Ley

SIGNAL TRANSDUCTION MEDIATED BY THE T CELL ANTIGEN x

323

P1: FRK

Annu. Rev. Immunol. 2002.20:709-760. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

February 1, 2002

12:47

Annual Reviews

AR152-FM

CONTENTS

xi

RECEPTOR: THE ROLE OF ADAPTER PROTEINS, Lawrence E. Samelson INTERACTION OF HEAT SHOCK PROTEINS WITH PEPTIDES AND ANTIGEN PRESENTING CELLS: CHAPERONING OF THE INNATE AND ADAPTIVE IMMUNE RESPONSES, Pramod Srivastava CHROMATIN STRUCTURE AND GENE REGULATION IN THE IMMUNE SYSTEM, Stephen T. Smale and Amanda G. Fisher PRODUCING NATURE’S GENE-CHIPS: THE GENERATION OF PEPTIDES FOR DISPLAY BY MHC CLASS I MOLECULES, Nilabh Shastri, Susan

371

Schwab, and Thomas Serwold

395 427

463

THE IMMUNOLOGY OF MUCOSAL MODELS OF INFLAMMATION, Warren Strober, Ivan J. Fuss, and Richard S. Blumberg T CELL MEMORY, Jonathan Sprent and Charles D. Surh

GENETIC DISSECTION OF IMMUNITY TO MYCOBACTERIA: THE HUMAN MODEL, Jean-Laurent Casanova and Laurent Abel ANTIGEN PRESENTATION AND T CELL STIMULATION BY DENDRITIC CELLS, Pierre Guermonprez, Jenny Valladeau, Laurence Zitvogel, Clotilde Th´ery, and Sebastian Amigorena

495 551 581

621

NEGATIVE REGULATION OF IMMUNORECEPTOR SIGNALING, Andr´e Veillette, Sylvain Latour, and Dominique Davidson

669

CPG MOTIFS IN BACTERIAL DNA AND THEIR IMMUNE EFFECTS, Arthur M. Krieg

PROTEIN KINASE Cθ

709 IN

T CELL ACTIVATION, Noah Isakov and Amnon

Altman

RANK-L AND RANK: T CELLS, BONE LOSS, AND MAMMALIAN EVOLUTION, Lars E. Theill, William J. Boyle, and Josef M. Penninger PHAGOCYTOSIS OF MICROBES: COMPLEXITY IN ACTION, David M. Underhill and Adrian Ozinsky

761 795 825

STRUCTURE AND FUNCTION OF NATURAL KILLER CELL RECEPTORS: MULTIPLE MOLECULAR SOLUTIONS TO SELF, NONSELF DISCRIMINATION, Kannan Natarajan, Nazzareno Dimasi, Jian Wang, Roy A. Mariuzza, and David H. Margulies

853

INDEXES Subject Index Cumulative Index of Contributing Authors, Volumes 1–20 Cumulative Index of Chapter Titles, Volumes 1–20

ERRATA An online log of corrections to Annual Review of Immunology chapters (if any, 1997 to the present) may be found at http://immunol.annualreviews.org/

887 915 925

8 Feb 2002

10:10

AR

AR152-23.tex

AR152-23.SGM

LaTeX2e(2001/05/10) P1: GJC 10.1146/annurev.immunol.20.120701.074321

Annu. Rev. Immunol. 2002. 20:761–94 DOI: 10.1146/annurev.immunol.20.120701.074321 c 2002 by Annual Reviews. All rights reserved Copyright °

PROTEIN KINASE Cθ

IN

T CELL ACTIVATION

Noah Isakov1 and Amnon Altman2 Annu. Rev. Immunol. 2002.20:761-794. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

1

Department of Microbiology and Immunology, Faculty of Health Sciences, and the Cancer Research Center, Ben Gurion University of the Negev, Beer Sheva 84105, Israel; e-mail: [email protected] 2 Division of Cell Biology, La Jolla Institute for Allergy and Immunology, 10355 Science Center Drive, San Diego, California 92121; e-mail: [email protected]

Key Words PKCθ, TCR/CD28, immune synapse, IL-2, NF-κB/AP-1 ■ Abstract The novel protein kinase C (PKC) isoform, PKCθ , is selectively expressed in T lymphocytes and is a sine qua non for T cell antigen receptor (TCR)triggered activation of mature T cells. Productive engagement of T cells by antigenpresenting cells (APCs) results in recruitment of PKCθ to the T cell–APC contact area—the immunological synapse—where it interacts with several signaling molecules to induce activation signals essential for productive T cell activation and IL-2 production. The transcription factors NF-κB and AP-1 are the primary physiological targets of PKCθ, and efficient activation of these transcription factors by PKCθ requires integration of TCR and CD28 costimulatory signals. PKCθ cooperates with the protein Ser/Thr phosphatase, calcineurin, in transducing signals leading to activation of JNK, NFAT, and the IL-2 gene. PKCθ also promotes T cell cycle progression and regulates programmed T cell death. The exact mode of regulation and immediate downstream substrates of PKCθ are still largely unknown. Identification of these molecules and determination of their mode of operation with respect to the function of PKCθ will provide essential information on the mechanism of T cell activation. The selective expression of PKCθ in T cells and its essential role in mature T cell activation establish it as an attractive drug target for immunosuppression in transplantation and autoimmune diseases.

INTRODUCTION An effective immune response depends on the ability of specialized immunocytes to identify foreign molecules and respond by differentiation into mature effector cells. This tightly regulated process is mediated by a cell-surface antigen recognition apparatus and a complex intracellular receptor-coupled signal-transducing machinery, which operate at high fidelity to discriminate self from nonself antigens. The T cell antigen-specific receptor (TCR) is composed of two covalently bound polymorphic subunits, which provide antigen specificity, in association with at least four different types of invariant chains that are essential for signal transduction. Activation of T lymphocytes requires sustained physical interaction of the TCR 0732-0582/02/0407-0761$14.00

761

24 Jan 2002

12:17

Annu. Rev. Immunol. 2002.20:761-794. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

762

AR

ISAKOV

AR152-23.tex

¥

AR152-23.SGM

LaTeX2e(2001/05/10)

P1: GJC

ALTMAN

with a major histocompatibility complex (MHC)–presented peptide antigen. This interaction results in a temporal and spatial reorganization of multiple cellular elements at the T cell–antigen presenting cell (APC) contact region, a specialized region referred to as the immunological synapse (IS) (1) or the supramolecular activation cluster (SMAC) (2; see below). TCR engagement by a peptide/MHC complex is one essential signal (signal 1) for T cell activation. However, productive T cell activation depends on an additional signal (signal 2) that can be provided by a number of costimulatory receptors. The major costimulatory signal for T cell activation is provided by interaction of the T cell surface molecule CD28 with its CD80/CD86 (B7-1/B7-2) ligands on APCs (3–6). This interaction plays an essential role in TCR-mediated IL-2 production. The nature of the CD28 costimulatory signal is relatively poorly defined. It may result from unique biochemical signals such as activation of phosphatidylinositol 3-kinase (PI3K) and/or its ability to enhance TCR signals in a relatively nonspecific manner, perhaps by facilitating formation of the IS and stabilizing it via its positive effect on lipid raft clustering (7). Studies during the last two decades revealed and characterized many types of effector molecules that play a role in the TCR-linked signal transduction machinery. This work has led to the recognition of the critical role of multiple enzymes and adapter proteins in the early signaling events downstream of the engaged TCR. During the last 12 years, much work was centered on the role of reversible tyrosine phosphorylation and the related enzymes, i.e., protein tyrosine kinases and phosphatases, in regulating T cell activation. However, historically, the earliest studies on T cell activation have focused on the regulatory role of inositol phospholipid turnover and the resulting second messengers. These early studies revealed that TCR engagement leads to phospholipase C-γ 1 (PLC γ 1)-mediated hydrolysis of membrane inositol phospholipids and subsequent production of inositol phosphates and diacylglycerol (DAG). These two second messengers stimulate, in turn, an elevation in intracellular Ca2+ concentration ([Ca2+]i) and activation of protein kinase C (PKC). Further support for the importance of these events in T cell activation was obtained using a combination of Ca2+ ionophore and phorbol ester tumor promoters (such as PMA) that mimicked TCR signals leading to full T cell activation resulting in IL-2 production and proliferation (8–11). The discovery of PKC as a lipid- and Ca2+-dependent serine/threonine kinase that serves as a cellular receptor for tumor-promoting phorbol ester (12) implicated PKC in activation and mitogenesis of T cells. This was further substantiated by demonstrating that cellular depletion of PKC by prolonged treatment with phorbol esters, or PKC inhibitory drugs, significantly inhibited T cell activation (13). During the same period, progress has been made in isolating the cDNAs encoding different PKC enzymes. This work revealed that PKC constitutes a family of multiple enzymes encoded by different genes that have distinct biochemical properties, expression profiles, and physiological functions. However, the specific contribution of distinct PKC isoforms to TCR-coupled signaling pathways has not been resolved until very recently. This gap in our knowledge was largely due to

8 Feb 2002

10:42

AR

AR152-23.tex

AR152-23.SGM

LaTeX2e(2001/05/10)

P1: GJC

PKCθ IN T CELL ACTIVATION

763

the relatively slow progress in identifying specific physiological substrates and functions of distinct PKC isoenzymes in T cells. Recent studies have begun to fill this gap by providing new information on PKCθ, which is now known to selectively mediate several essential functions in TCR-linked signaling leading to cell activation, differentiation, and survival (Figure 1).

Annu. Rev. Immunol. 2002.20:761-794. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

ISOLATION OF PKCθ The discovery of PKC and its initial isolation and characterization revealed that this enzyme family is ubiquitously expressed and is also abundant in lymphoid tissues, peripheral blood mononuclear cells, and leukemic cell lines (14, 15). Later studies demonstrated that PKC is the cellular receptor for phorbol ester tumor promoters (16–18), which provided a long sought-after explanation for the observed effects of PMA on T cell mitogenesis (19, 20). In addition, both PKC-interacting phorbol esters and T cell proliferation–inducing agents (such as mitogenic lectins or antireceptor antibodies) induced a similar redistribution of PKC in T cells, characterized by its translocation from the cytosol to the particulate fraction (21–27). Together with the observation that PMA binding upregulates PKC activity, these results were consistent with the idea that PKC plays a key role in the activation response of T cells. Cloning of several PKC genes led to the realization that these enzymes represent a large family encoded by distinct genes. Analysis of their tissue distribution demonstrated that different isoenzymes are expressed within individual cells and that many isoenzymes are expressed in a wide range of cell types and tissues. The finding that PKC activity is essential for TCR/CD3-induced T cell activation (13) led us to initiate a search for PKC isoforms that may play a specific role in T cell development and/or activation. These efforts led to the identification of a new member of the Ca2+-independent novel PKC subfamily, termed PKCθ (28). Other investigators also cloned the corresponding mouse and human cDNAs (29, 30). Chromosomal mapping located the human PKCθ gene to the short arm of chromosome 10 (10p15) (31), a region prone to mutations that lead to T cell leukemia and lymphomas and T cell immunodeficiencies (32, 33).

PKCθ PROTEIN STRUCTURE The predicted structure of PKCθ displays the highest homology with members of the Ca2+-independent novel PKC subfamily, including PKCδ, ε, and η, and a functional assay confirmed dependency of its catalytic activity on phospholipids but not on Ca2+ ions (34). Within this subfamily, PKCθ is most highly related to PKCδ because the V1 domains of these two enzymes share a 49% homology (28). The amino terminus of PKCθ includes a negative regulatory pseudosubstrate region

24 Jan 2002

12:17

Annu. Rev. Immunol. 2002.20:761-794. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

764

AR

ISAKOV

AR152-23.tex

¥

AR152-23.SGM

LaTeX2e(2001/05/10)

P1: GJC

ALTMAN

that, under resting conditions, interacts in cis with the substrate-binding region in the catalytic domain. The resulting closed conformation of the enzyme leaves the catalytic domain inaccessible to substrate (35). These intrinsic regulatory properties of the enzyme were the basis for the construction of genetically modified PKCθ forms for functional studies. Thus, a constitutively active PKCθ was generated by replacing a critical Ala (A148) by Glu in the PKCθ pseudosubstrate region (36, 37). This introduces a negatively charged amino acid that mimics a phosphate group and creates repelling forces between the pseudosubstrate and the substrate-binding region. The resulting change in the conformation leaves the catalytic site free for interaction with substrates, thereby converting the enzyme into a constitutively active kinase (PKCθ-A/E). Conversely, replacement of a critical lysine (Lys409) by an arginine in the ATP-binding region results in the formation of a molecule (PKCθ-K/R) with an intact ability to bind substrates but an inability to phosphorylate them (36, 37). This PKCθ mutant can function in vivo as a dominant-negative protein that blocks or reduces physiological effects mediated by the endogenous wild-type PKCθ. Both PKCθ mutants were extremely useful in transfection studies and contributed significantly to the functional studies of this enzyme.

TISSUE DISTRIBUTION OF PKCθ Analysis of PKCθ distribution in mouse and rat tissues demonstrated a relatively restricted pattern of expression (28–30, 34, 38–42). High levels of PKCθ were found in skeletal muscle and lymphoid tissues, predominantly in the thymus and lymph nodes, with lower levels in spleen, and undetectable expression in the bone marrow (43). Expression in the muscle was restricted to the plasma membrane, and immunochemical analysis demonstrated association of PKCθ with the sarcolemma of skeletal muscle and its localization in the neuromuscular junction (40). Analysis of PKCθ mRNA expression during mouse development by in situ hybridization revealed expression in yolk sac blood islands and in the liver, and later in the thymus and skeletal muscle. In addition, high expression was detected in the embryonic nervous system, including spinal ganglia, spinal cord, trigeminal and facial ganglia, and a subsection of the thalamus (42). The highest expression of PKCθ mRNA in the adult thymus was found in the cortex (42). Further analysis of different lymphoid and myeloid cell types demonstrated selective expression of PKCθ in T, but not B, lymphocytes and in platelets but not erythrocytes, polymorphonuclear neutrophils, monocytes, or macrophages (44). Among T cells, CD4+ and CD8+ single positive peripheral blood T cells, and also CD4+CD8+ double positive thymocytes, were found to express high levels of the PKCθ protein (43). Studies in leukemic cell lines representing distinct stages of differentiation suggested that PKCθ might be developmentally regulated (44). Since cord blood CD34+ stem cells express relatively low levels of PKCθ (N. Isakov, unpublished observation), it is possible that stem cell differentiation into mature T cells progresses along a pathway involving upregulation of PKCθ, while a concomitant loss of PKCθ occurs upon progression to mature B cells.

24 Jan 2002

12:17

AR

AR152-23.tex

AR152-23.SGM

LaTeX2e(2001/05/10)

P1: GJC

PKCθ IN T CELL ACTIVATION

765

The potential role of PKCθ in hematopoiesis and in cell commitment and differentiation into a specific lineage awaits further genetic studies in PKCθ gene transgenic mice or targeted gene disruption models, which may provide more conclusive information. This analysis may also provide clues for the function of PKCθ in muscle and nerve cells, which currently remains enigmatic.

Annu. Rev. Immunol. 2002.20:761-794. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

PKCθ : A MASTER INDUCER OF T CELL TRANSCRIPTION FACTORS AND IL-2 Findings that PKCθ plays a role in the induction of early activation genes and regulation of their corresponding transcription factors provided the first clue that PKCθ is intimately involved in T cell activation. Two major groups of transcription factors were tested in this respect. The first group includes factors that are induced by phorbol esters and regulate gene transcription from TPA response element (TRE)-containing promoters. The second group includes factors linked to TCR/CD28-coupled signal transduction pathways that are essential for transcriptional regulation of T cell growth-promoting genes, predominantly the IL-2 gene.

The JNK/AP-1 Pathway The AP-1 transcription factor consists of a dimer of Jun and/or Fos proteins that regulate transcription from multiple TRE-containing genes (45, 46), including the IL-2 gene (47, 48). The ability of PKC-activating phorbol esters to stimulate AP-1 implicated PKC in AP-1 activation. In an attempt to define a role for PKCθ in T cell activation, we initially turned our attention to the potential function of PKCθ as an upstream regulator of AP-1. Stable overexpression of either PKCθ or PKCα in Jurkat T cells, followed by stimulation with an anti-CD3 antibody plus PMA, promoted a ∼sixfold increase in IL-2 production (36). However, when this analysis was extended to reporter genes driven by distinct binding sites in the IL-2 promoter, only PKCθ, but not PKCα, activated AP-1 (36). Constitutively active PKCθ stimulated AP-1 activity in resting T cells, and, conversely, dominant negative PKCθ blocked the PMA-induced AP-1 activity (36). These results indicated that PKCθ is an important component of the signaling pathway leading to AP-1 activation in T cells. This was in full agreement with the later finding that PKCθ -deficient primary T cells display a defect in receptor-stimulated AP-1 function (49). Regulation of AP-1 activity is a complex process mediated by both transcriptional and posttranscriptional events (46), and it involves the activation of the c-Jun N-terminal kinase (JNK), which selectively phosphorylates two positive regulatory serine residues in the c-Jun activation domain, i.e., Ser-63 and -73 (46). T cells display a unique two-signal requirement for JNK activation that involves a combination of TCR and CD28 signals, which can be mimicked by phorbol ester plus Ca2+ ionophore (50). This finding suggested a role for PKC in JNK activation in T cells and, given the selective role of PKCθ in AP-1 stimulation (36), raised the possibility that PKCθ has a specialized role in JNK activation. Several groups

24 Jan 2002

12:17

Annu. Rev. Immunol. 2002.20:761-794. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

766

AR

ISAKOV

AR152-23.tex

¥

AR152-23.SGM

LaTeX2e(2001/05/10)

P1: GJC

ALTMAN

independently demonstrated that PKCθ , but not other PKC isoforms, activates JNK in a T cell–specific manner and, furthermore, synergizes with calcineurin in inducing this effect (51–53). In addition, PKCθ activated SEK1 (MKK4), the immediate MAP kinase that phosphorylates and activates JNK (51). However, the immediate target for PKCθ -mediated phosphorylation in the JNK pathway is unknown. In contrast to the selective effect of PKCθ on JNK, multiple PKC isoforms nonselectively activated ERK (52), a distinct MAP kinase operating along a parallel signaling pathway, which is linked to different cell surface receptors or is activated under different stimulation conditions. Although there is little doubt that PKCθ selectively stimulates SEK1 and JNK activity in T cells, the physiological relevance of this pathway is less clear because peripheral T cells from PKCθ -deficient mice display intact TCR/CD28-induced JNK activation in the face of defective AP-1 activation (49). These findings suggest not only that PKCθ plays an important and nonredundant role in AP-1 activation, but also that its function in JNK activation can be compensated by another PKC isoform or even by a PKC-independent pathway. It is possible, however, that nonphysiological compensatory mechanisms are upregulated in these mice, given the chronic lack of PKCθ expression. Another possibility is that the target of PKCθ in the pathway leading to AP-1 activation is not JNK. In this regard, it would be interesting to determine whether PKCθ positively regulates a pathway involved in the induction of c-Fos. The role of JNK in IL-2 production was questioned when T cells from mice genetically deficient in both JNK1 and JNK2 were found to display enhanced IL-2 production in response to combined TCR/CD28 stimulation (54). Furthermore, na¨ıve T cells do not express detectable levels of JNK. However, JNK expression was highly upregulated upon primary stimulation of these cells (55). These findings raise the possibility that selective activation of JNK by PKCθ plays a more important role in memory or effector T cells.

NF-κB Activation NF-κB is an inducible heterodimeric transcription factor involved in activation of a large number of genes in response to stress conditions such as infection or inflammation that require rapid reprogramming of gene expression. NF-κB is activated by a wide range of external stimuli, which, in T lymphocytes, includes cytokines, mitogens, and TCR/CD28 costimulation (56). NF-κB is maintained in an inactive form in the cytoplasm of resting cells; upon cell stimulation, it translocates to the nucleus where it regulates gene transcription. Sequestration of NF-κB in the cytosol is regulated by one of several inhibitory proteins, the IκBS, which interact with NF-κB and mask its nuclear localization signal. NF-κB activation by proinflammatory cytokines, such as TNF-α and IL-1β, has been thoroughly studied and was found to involve a series of signaling intermediates that lead to activation of IκB kinase (IKK), upstream kinases such as NF-κB– inducing kinase (NIK) (57, 58), AKT/PKB (59), MEKK1 (60), or other regulatory proteins (61). One or more of these kinases then phosphorylates and activates

24 Jan 2002

12:17

AR

AR152-23.tex

AR152-23.SGM

LaTeX2e(2001/05/10)

P1: GJC

Annu. Rev. Immunol. 2002.20:761-794. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

PKCθ IN T CELL ACTIVATION

767

IKKα and IKKβ, the two catalytic subunits of the IKK complex (62–64), which are able to phosphorylate two critical regulatory serine residues in the IκB amino terminus (65, 66). The resulting conformational change in IκB creates a novel ubiquitin-ligase recognition site, which triggers IκB ubiquitination, followed by its degradation in the proteasome (61, 67, 68). Elimination of IκB unmasks the nuclear localization signal in NFκB, which is then translocated to the nucleus to promote transcription of NFκB-responsive target genes (69). Activation of NF-κB via the TCR/CD28 receptor–coupled signaling pathway (in contrast to the TNF-α-R- or IL-1β-R–coupled pathway) is mediated by different signaling intermediates, and so far it is less well defined. Although TCR stimulation is sufficient to activate NFAT (70), activation of NF-κB requires simultaneous costimulation of the CD28 receptor (71–73). The transcriptional element, which serves as a target for TCR/CD28 costimulation, is the CD28 response element (CD28RE) in the IL-2 promoter (47). CD28RE is a combinatorial binding site for NF-κB and AP-1 (74, 75). TCR/CD28 costimulation was found to induce phosphorylation and activation of IKKs (76), possibly by upstream kinases such as NIK, Cot, or MEKK1 (77, 78). The immediate upstream regulator of IKKs in this T cell activation pathway has not yet been defined. However, recent studies demonstrated that PKCθ plays a role in the TCR/CD28-coupled signaling pathway, which leads to phosphorylation and activation of IKKs (79, 80). Overexpression of a constitutively active form of PKCθ , but not other isoforms (such as PKCα, ε, η, and ζ ), activated both NF-κB and CD28RE reporter genes in Jurkat T cells (79–81). This effect was T cell specific because PKCθ was relatively inefficient in activation of NF-κB in nonlymphoid 293T cells. Conversely, a kinase-deficient PKCθ or a PKCθ antisense vector inhibited TCR/CD28 costimulation and NF-κB activation in Jurkat cells (80), and a relatively selective PKCθ inhibitor (rottlerin) blocked the TCR/CD28-induced CD28RE activation and NF-κB translocation to the nucleus (79). Furthermore, rottlerin blocked the TCR/CD28-induced activation of IKK in primary human CD4+ T, in which an activation-dependent physical interaction between PKCθ and the IKK complex was also observed (81). Inhibition of NF-κB activation in Jurkat cells by a kinase-deficient PKCθ mutant or by rottlerin was selective for the TCR/CD28 costimulation-coupled signaling pathway and did not affect NF-κB activation induced by TNF-α (79, 80). PKCθ was unable to replace the TCR or the CD28 signals required for NF-κB activation because enhanced activation of the CD28RE mediated by transient overexpression of wild-type PKCθ was strictly dependent on TCR/CD28 costimulation (79). This is in accordance with the findings that CD28 costimulation further augmented the TCR-induced membrane translocation and enzymatic activation of PKCθ (79), as well as the recruitment and localization of PKCθ to membrane lipid rafts (see below). PKCθ-induced activation of NF-κB is accompanied by activation of IKKβ but not IKKα (79, 80). However, this is most likely not a direct effect, and the kinase that directly phosphorylates and activates IKKβ in this PKCθ -dependent pathway has not been identified. Activated IKKβ phosphorylates IκB, thereby promoting

24 Jan 2002

12:17

768

AR

ISAKOV

AR152-23.tex

¥

AR152-23.SGM

LaTeX2e(2001/05/10)

P1: GJC

ALTMAN

Annu. Rev. Immunol. 2002.20:761-794. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

its ubiquitination and proteasome-dependent degradation—an effect that could be blocked by the proteasome-specific inhibitor, MG132 (79). The essential and selective role of PKCθ in NF-κB activation is evident from the recent finding that PKCθ-deficient T cells display a severe defect in receptorinduced NF-κB activation (49). When contrasted with the observation that JNK1/ JNK2-deficient primary T cells display intact IL-2 production (54), these findings implicate the NF-κB cascade, rather than JNK, as the major and physiologically most critical selective target of PKCθ in the TCR/CD28 costimulatory pathway that leads to IL-2 production, at least in the case of primary T cells.

The Role of PKCθ in Induction of the IL-2 Gene Productive T cell activation leads to the synthesis and secretion of multiple cytokines, including IL-2, which functions as a major growth-promoting factor for lymphocytes. The IL-2 gene promoter possesses consensus binding sites for several known transcription factors, including AP-1, NF-κB, NFAT, and Oct, and additional regulatory sites, such as the CD28RE (47, 48). A cooperative interaction between these factors is required for efficient induction of the IL-2 gene (47, 48). As noted earlier, the TCR-coupled signaling events leading to IL-2 gene transcription can be mimicked by a combination of phorbol ester and Ca2+ ionophore. The target for the stimulatory effect of Ca2+ ionophore and the ensuing increase in [Ca2+]i was subsequently identified as calcineurin, a Ca2+-dependent Ser/Thr phosphatase that dephosphorylates cytosolic NFAT, thus enabling its nuclear translocation and transcriptional activation (47, 48, 82). The finding that JNK activation depends on the same two signals (50), coupled with the selective ability of PKCθ to activate JNK (42, 51–53), suggested that PKCθ is a major target for the PMA signal and that it plays an important rate-limiting role in IL-2 induction. Several laboratories independently reported that transfecting T cells with a combination of constitutively active PKCθ and calcineurin plasmids induced synergistic activation of the IL-2 gene (52, 53). Conversely, a dominant-negative PKCθ mutant selectively inhibited activation of the IL-2 promoter. The requirement for PKCθ was specific because no other PKC isoenzymes reproduced these effects, which were mainly attributed to NFAT activation (52). The role of PKCθ as a selective upstream regulator of NFAT was further substantiated in Nef-expressing cells, in which Nef-induced NFAT-dependent gene transcription was found to require PKCθ (83). The important role of PKCθ in IL-2 gene regulation is not surprising in view of the fact that the distal NFAT site in the IL-2 promoter cooperatively binds NFAT and AP-1 proteins (48, 82), which, in turn, are activated by calcineurin and PKCθ, respectively. These studies strongly suggest that PKCθ is the immediate target for the pharmacological effect of PMA and thus represents the long sought-after second TCR signal required for IL-2 induction. While the essential role of PKCθ in the regulation of IL-2 gene transcription is well established, the relevant PKCθ downstream signaling events and the PKCθ -responsive transcription factors have been only partially characterized.

24 Jan 2002

12:17

AR

AR152-23.tex

AR152-23.SGM

LaTeX2e(2001/05/10)

P1: GJC

Annu. Rev. Immunol. 2002.20:761-794. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

PKCθ IN T CELL ACTIVATION

769

Besides AP-1 and NF-κB activation, PKCθ appears to regulate IL-2 gene transcription in a different, potentially negative manner. This regulation involves a site in the IL-2 promoter located about −180 bp from the transcriptional start site. This site was found to bind complexes of cAMP response element–binding protein (CREB)/cAMP response element modulator (CREM), ATF2/c-Jun, and Jun-Jun/Oct complexes (84). However, induction of T cell anergy led to an increase in binding of only the CREB/CREM complex. Moreover, mutation of the CREB/CREM-binding site, but not of the Jun-Jun/Oct site, reduced the susceptibility of T cells to anergy induction (84). These findings suggest that the −180 region of the IL-2 promoter is the target of a CREB/CREM transcriptional inhibitor that contributes to the repression of IL-2 production in T cell anergy (5). A more recent study demonstrated that binding of CREB to this site is specifically regulated by PKCθ (85). T cell activation induced a PKCθ-mediated phosphorylation of CREB, and TCR/CD28 costimulation or PMA plus Ca2+ ionophore– induced binding of phospho-CREB to the −180 site in the IL-2 promoter (85). Rottlerin decreased CREB phosphorylation levels and CREB binding to the IL-2 promoter. PKCθ significantly increased the activity of a reporter gene driven by two tandem copies of the IL-2 promoter −180 site. PKCθ also increased the reporter activity driven by the −575/+57 region of the IL-2 promoter/enhancer, but activity was significantly reduced when two nucleotides in the IL-2 promoter CREB-binding site were mutated (85). The biological implications of this regulatory event were not addressed in this study. However, given the finding that anergic T cells display increased binding of CREB to the same site (84), these findings raise the possibility that PKCθ -mediated CREB phosphorylation negatively regulates IL-2 transcription, thereby driving the responding T cells into an anergic state. However, it is difficult to reconcile such a negative regulatory role of PKCθ with the findings that PKCθ is required for mature T cell activation (49, 86). It is possible that this represents a feedback regulatory mechanism to terminate IL-2 production by activated T cells. Alternatively, the increased CREB phosphorylation observed in anergic T cells may be mediated by another Ser/Thr kinase that antagonizes the activating function of PKCθ and, therefore, induces anergy. Mapping of the phosphorylation sites in CREB and identification of the relevant phosphorylating kinases should resolve these questions. A different approach to analyzing PKC isoenzyme-specific functions in peripheral blood T cells was adopted by Szamel et al., who introduced PKC-specific neutralizing antibodies into cells by electroporation. They found that PKCθ was essential for IL-2 receptor expression, but not for IL-2 synthesis in activated T cells (87). However, the second finding is not supported by the phenotype of PKCθ gene knockout mice, whose T cells display defective IL-2 production (49). This discrepancy may reflect lack of specificity of the neutralizing antibodies. A recent study demonstrated a potential role for PKCθ in IL-4 production (88), but the contribution of PKCθ to the activation of other T cell cytokine genes is an open question. The critical role of PKCθ in induction of IL-2 production and the concomitant T cell proliferation was confirmed by the findings that mature T cells of PKCθ

24 Jan 2002

12:17

770

AR

ISAKOV

AR152-23.tex

¥

AR152-23.SGM

LaTeX2e(2001/05/10)

P1: GJC

ALTMAN

knockout mice produced very little IL-2 and underwent minimal proliferation in response to TCR/CD28 costimulation (49). It is surprising, however, that T cell development and the number and phenotype of T cells in the peripheral blood of PKCθ knockout mice were unaltered.

Annu. Rev. Immunol. 2002.20:761-794. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

A Link Between PKCθ and T Cell Anergy? When a TCR signal is delivered in the absence of CD28 costimulation, the T cell enters a stable state of unresponsiveness, termed anergy. Thus, a CD28 signal prevents anergy induced by signal 1 alone (3, 5, 6). Therefore, the CD28/B7 pathway is an attractive target for developing drugs that will either stimulate this pathway to augment antigen-specific T cell activation, e.g., in tumor-specific responses, or suppress it, e.g., in autoimmune and GvH diseases and in organ transplantation. Indeed, agonistic or antagonistic modulation of CD28 costimulation has profound effects on these disease conditions in animal models (4). Considerable effort has focused on analyzing the molecular basis of the role of CD28 in T cell anergy. T cell anergy is associated with defective IL-2 production, and addition of exogenous IL-2 reverses the anergic state (5). Anergy reflects, not a global failure of TCR signaling pathways, but rather a selective defect in activation of a subset of signaling pathways normally induced by TCR agonists. This is substantiated by findings that NFAT1 activation and Ca2+ mobilization remain intact in anergic T cells (89–91). Past studies revealed defects in the TCR signaling pathway of anergic T cells at several levels: TCR tyrosine phosphorylation and expression or activity of the Src-family kinases Lck and Fyn (92–95); Ras and MAP kinase activation (90, 96, 97); excessive activation of the negative regulatory small GTPase, Rap1 (98); stimulation of an inhibitory Cbl-dependent pathway (94, 95); or induction of the transcription factors AP-1 (89, 91, 99) and NF-κB (99). It is of particular interest that activation of the two latter transcription factors is absolutely dependent on PKCθ in mature T cells (49), and they both bind to the CD28RE (74, 75), which is the major target for TCR/CD28 costimulation (100). Indeed, increased AP-1 expression confers resistance to anergy induction in Agspecific T cells (101), and overexpression of c-Jun and p65 (a member of the NF-κB family) substitutes for B7 costimulation by targeting the CD28RE (102). Because PKCθ is required for activation of the same two transcription factors, it is tempting to propose that PKCθ is a target and transducer of TCR/CD28 costimulatory signals important for preventing anergy. It would, therefore, be of interest to determine whether experimental strategies that specifically block PKCθ function induce an anergy-like state in antigen-specific T cells.

PKCθ AND THE T CELL SYNAPSE T cell activation requires sustained TCR interaction with MHC-bound peptide antigen at the T cell–APC contact region. Productive interaction results in biochemical changes and reorganization of specific membrane domains, which lead

24 Jan 2002

12:17

AR

AR152-23.tex

AR152-23.SGM

LaTeX2e(2001/05/10)

P1: GJC

Annu. Rev. Immunol. 2002.20:761-794. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

PKCθ IN T CELL ACTIVATION

771

to the formation of a highly ordered signaling complex at the contact site, the so-called IS (1), a term originally coined in reference to the polarized cytokine secretion by antigen-stimulated T cells (103). Formation of a functional IS also involves (a) the assembly of signaling complexes consisting of TCRs, costimulatory accessory receptors (such as CD28, CD4/CD8, or LFA-1), and intracellular signaling effector proteins (1, 2, 104); (b) reorganization of the actin cytoskeleton (104–106); and (c) clustering of specialized membrane microdomains or lipid rafts (107–110). A more detailed analysis of the T cell–APC contact region revealed compartmentalization of molecules in at least two distinct identifiable areas of the synapse, the so-called central SMAC (cSMAC) and peripheral SMAC (pSMAC) (2). While the cSMAC is characterized by clustering of TCR and MHC molecules on the T cell and APC surfaces, respectively, the pSMAC in these two cell types is enriched with LFA-1 integrins and their ICAM-1 counter-receptors, respectively. The spatial organization and stability (or duration) of the IS determine the functional outcome of TCR engagement and underlie the fundamental phenomenon of differential T cell signaling (111). The initial findings linking PKCθ to the IS demonstrated that engagement of antigen-specific T cells by APCs led to a rapid, stable, and high-stoichiometry localization of PKCθ, but not other T cell–expressed PKCs (βI, δ, ε, η, and ζ ), to the T cell–APC contact site (112) and, more specifically, to the cSMAC (2). This clustering correlated with the catalytic activation of PKCθ, and it only occurred upon productive activation of T cells, i.e., upon exposure to APCs that were fed with optimal antigen concentrations leading to efficient proliferation. In contrast, altered peptide ligands or low peptide concentrations that induced weak or no detectable proliferation did not promote PKCθ recruitment to the cSMAC (112). Coclustering of talin and tubulin and formation and reorientation of the microtubule-organizing center (MTOC) were also observed under these conditions. Subsequently, it became clear that signaling molecules on the inner side of the cell membrane also segregate into two non-overlapping regions characterized by PKCθ and Lck at the cSMAC, just below the TCR, and talin molecules in the peripheral zone, where they can directly interact with the LFA-1 cytoplasmic tail (1, 2). A recent study addressed the role of CD28 versus LFA-1 in IS assembly and PKCθ recruitment to the cSMAC (C. E. Sedwick, K. Blaine, and J. Miller, submitted). While signaling through the TCR plus either LFA-1 or CD28 could recruit PKCθ to the IS, CD28 was required to localize PKCθ specifically to the cSMAC and to activate NF-κB. These results are consistent with our earlier findings that CD28 provides unique and essential signals required for the proper localization and activation of PKCθ .

PKCθ AND LIPID RAFTS The plasma membrane of many cell types, including T cells, contains glycosphingolipid-enriched membrane microdomains (GEMs) or detergent-insoluble glycolipid (DIG) fractions, which are enriched in multiple glycosyl-phosphatidylinositol

24 Jan 2002

12:17

Annu. Rev. Immunol. 2002.20:761-794. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

772

AR

ISAKOV

AR152-23.tex

¥

AR152-23.SGM

LaTeX2e(2001/05/10)

P1: GJC

ALTMAN

(GPI)-anchored proteins. Because of their distinct biochemical and physical properties, these microdomains are relatively resistant to non-ionic detergents, characterized by low buoyant density, and therefore can be isolated by sucrose gradient centrifugation. They can also be identified in vitro or in intact cells by their selective ability to associate with cholera toxin B (Ctx) subunit, which binds selectively to the lipid raft–enriched glycosphingolipid, GM1. These membrane microdomains are likely to correspond to the lipid rafts, which function as “floating” platforms for the assembly of signaling complexes, following the engagement of specific cellsurface receptors (114, 115). Distinct receptors or intracellular signaling molecules associate with lipid rafts in T cells either constitutively or following T cell activation, and lipid raft integrity is important for productive T cell activation (107–110). A recent study addressed the relationship between lipid rafts and the IS with respect to the subcellular localization of PKCθ (116). Anti-CD3 stimulation induced recruitment of a small amount of PKCθ to lipid rafts, an effect that was augmented by CD28 costimulation (116). Moreover, TCR/CD28 costimulation of primary human T cells induced a simultaneous translocation of endogenous IKK to the lipid rafts, where physical interaction with PKCθ was observed (81). TCR/CD28 cross-linking also increased the enzymatic activity of the lipid raft–residing PKCθ (116) and its ability to phosphorylate and activate IKKβ (81). Stimulation of antigen-specific T cells by a peptide antigen presented by syngeneic APCs induced simultaneous translocation and colocalization of PKCθ and Ctx-labeled membrane rafts at the T cell synapse (116). Overexpression of truncated versions of PKCθ revealed that its translocation to rafts required the N-terminal regulatory domain. Nevertheless, the presence of functional catalytic domain of PKCθ in the lipid rafts was essential for the activation response. Thus, replacement of the PKCθ regulatory domain by a short sequence (seven amino acids) derived from the Lck amino terminus (which includes a membrane targeting sequence with myristoylation and acylation sites) allowed the localization of the chimeric PKCθ molecule to the rafts of nonstimulated T cells. Lipid raft localization of this PKCθ form was, however, insufficient for induction of T cell activation (as assessed by activation of NF-κB), which occurred only following cell stimulation with anti-CD28 antibodies plus PMA (116). Previous studies demonstrated that PKCθ is a target for phosphorylation by Lck in activated T cells (117). Because both PKCθ and Lck localize to membrane rafts and are essential for the activation response, it was of interest to analyze whether Lck also affects PKCθ recruitment to membrane rafts. By comparing PKCθ cellular redistribution in activated wild-type versus Lck-deficient Jurkat T cells, Bi et al. found that Lck is essential for the inducible translocation of PKCθ to membrane rafts (116). Clustering of PKCθ in the rafts was also dependent on the enzymatic activity of Lck and could be inhibited by the Lck-specific inhibitor, PP2. Furthermore, the lipid raft–residing PKCθ was associated with Lck, as observed by coimmunoprecipitation studies. In addition, the Lck-mediated tyrosine phosphorylation of PKCθ was dependent on the integrity of lipid rafts and was necessary for optimal induction of PKCθ -dependent biological activities, such as NF-κB activation. However, it is possible that additional raft-residing effector

24 Jan 2002

12:17

AR

AR152-23.tex

AR152-23.SGM

LaTeX2e(2001/05/10)

P1: GJC

PKCθ IN T CELL ACTIVATION

773

molecules (rather than Lck only), which are involved in the T cell–activation process, also regulate PKCθ activity and distribution. This notion is supported by recent findings that intact ZAP-70 and SLP-76 are essential for activation of both PKCθ and its downstream effector molecule, NF-κB (118), and, furthermore, that PKCθ cooperates with Akt/PKB to activate NF-κB in T cells (119, 120).

Annu. Rev. Immunol. 2002.20:761-794. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

REGULATION OF PKCθ BY THE VAV/RAC PATHWAY Cytoskeletal components play critical roles in the regulation of T cell–APC contact and ensuing events leading to T cell activation. The actin cytoskeleton drives the assembly of large clusters of signaling molecules and formation of the IS (104, 105, 121). Localization of PKCθ to the synapse and its selective involvement in early TCR/CD28 signaling events leading to T cell activation and proliferation raise questions regarding the mechanism that enables the selective recruitment of PKCθ to the T cell synapse. Unlike other T cell–expressed PKC enzymes, a fraction of cellular PKCθ associates with the cytoskeleton upon T cell activation (37, 122). In light of recent findings that the Vav/Rac pathway plays an important role in reorganization of the T cell actin cytoskeleton and in TCR capping (123, 124), Villalba et al. studied the potential role of the Vav/Rac pathway in regulating the recruitment of PKCθ to the membrane and its activation. Vav was found to promote the translocation of PKCθ from the cytosol to the membrane and cytoskeleton (122). It also induced PKCθ activation in a CD3/CD28 costimulation pathway that was dependent on Rac and actin cytoskeleton reorganization. In addition, a TCR/CD28-coupled Vav signaling pathway that mediated the activation of JNK and the IL-2 gene, and upregulated CD69 expression, was dependent on intact PKCθ function because all of these Vav-induced responses were inhibited by a dominant-negative PKCθ mutant or by a selective PKCθ inhibitor (122). These findings reveal that the Vav/Rac pathway promotes the recruitment of PKCθ to the T cell synapse and its activation. However, the identity of the protein (or lipid?) that actually carries PKCθ to the T cell synapse remains unknown. It is possible that PKCθ specifically binds to a cytoskeletal protein or to some scaffold protein that is associated with a component of the T cell cytoskeleton or synapse. In this respect it is important to note that PKCθ was identified as the major Ser/Thr kinase that phosphorylates moesin, a member of the ezrin-radixin-moesin (ERM) family of membrane cytoskeleton–linking proteins (125). Phosphorylation occurred on Thr558, the in vivo phosphorylation site of moesin within its conserved actin-binding domain, which suggests that PKCθ may also regulate actin-based organization of membrane components via its effect on moesin. This effect may be relevant to synapse formation in APC-engaged T cells. However, overexpression of active PKCθ alone did not induce significant actin polymerization in T cells (122). Although chemokine stimulation, which induces T cell polarization, causes moesin to translocate to the uropod, i.e., the T cell pole opposite the IS (126), the localization of moesin in antigen-stimulated T cells has not been examined in detail.

24 Jan 2002

12:17

774

AR

ISAKOV

AR152-23.tex

¥

AR152-23.SGM

LaTeX2e(2001/05/10)

P1: GJC

ALTMAN

Annu. Rev. Immunol. 2002.20:761-794. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

Recent studies demonstrated that the CD47 cell surface receptor can synergize with the TCR in induction of early activation events, such as an intracellular [Ca2+]i rise (127). Ligation of CD47 induced actin polymerization and association of PKCθ with the cytoskeleton—effects that were further augmented by simultaneous coligation of the TCR. Signaling by CD47 required its targeting to membrane rafts. Nevertheless, the synergy between CD47 and the TCR occurred via a unique mechanism distinct from that of CD28-induced costimulation, which had no effect on PKCθ association with the cytoskeleton (127).

A UNIQUE, PI3K-DEPENDENT MECHANISM OF PKCθ ACTIVATION PKCθ displays several unique properties that distinguish it from other T cell– expressed PKCs. First, PKCθ is selectively regulated by a Vav/Rac pathway and a fraction of it is associated with the actin cytoskeleton (122). Second, antigen stimulation induces recruitment of PKCθ, but not other PKCs, to the cSMAC in the IS (2, 112). These distinctive features suggest that a unique mechanism, other than (or in addition to) the conventional DAG- and PLC-dependent pathway for PKC activation, regulates PKCθ activation because PLCγ 1-induced activation of PKC would also be expected to activate other conventional and novel PKC isoforms. The nature of this putative mechanism was addressed in a recent study (128). Using three independent approaches, i.e., a selective PLC inhibitor, a PLCγ 1deficient T cell line, or a dominant-negative PLCγ 1 mutant, we demonstrated that CD3/CD28-induced membrane recruitment and activation of PKCθ are PLC independent. In contrast, the same inhibitory strategies blocked the membrane translocation of PKCα. On the other hand, membrane or lipid raft recruitment of PKCθ was absent in T cells treated with PI3K inhibitors or in Vav-deficient T cells and was enhanced by constitutively active PI3K (128; M. Villalba, K. Bi, P. Bushway, E. Reits, J. Neefjes, G. Baier, R. T. Abraham, and A. Altman, submitted). These findings do not completely rule out a requirement for DAG binding to the PKCθ C1 domain as an important step in its membrane binding and activation. Some DAG may still be required to initiate PKCθ membrane binding. Although this event per se may not be sufficient to recruit PKCθ to specific membrane compartments such as the IS or lipid rafts, it may facilitate its interaction with membrane or cytoskeletal components required for translocation of PKCθ to the cSMAC and its full activation. Such an interacting partner could be a membranelocalized protein kinase that transphosphorylates PKCθ or an adapter/scaffold protein that recruits and anchors it to specific membrane microdomains (2, 112) or lipid rafts (116). The requirement for both Vav and PI3K in PKCθ membrane translocation and activation most likely reflects the fact that Vav is a critical target for PI3K in a single pathway regulating PKCθ activation in the IS. This putative mechanism is consistent with the finding that Vav is activated by PIP3 binding to its PH domain (129), a process that promotes Vav recruitment to the membrane and facilitates its

24 Jan 2002

12:17

AR

AR152-23.tex

AR152-23.SGM

LaTeX2e(2001/05/10)

P1: GJC

PKCθ IN T CELL ACTIVATION

775

Annu. Rev. Immunol. 2002.20:761-794. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

catalytic activation by regulatory tyrosine phosphorylation (130). The requirement of both TCR and CD28 signals for stable activation and membrane or lipid raft translocation of PKCθ (79, 116) could reflect this dual regulatory mechanism for Vav activation. The finding that Vav and constitutively active PI3K do not cooperate to enhance membrane translocation of PKCθ (128) is also consistent with the notion that PI3K and Vav function in a single pathway. However, the possibility that Vav and PI3K function in two independent pathways to promote PKCθ translocation and activation cannot be formally ruled out.

RELATIONSHIP BETWEEN PKCθ AND THE RAS PATHWAY Ras activation is an essential component of TCR-initiated T cell activation leading to IL-2 production. Ras can be activated by a tyrosine kinase–dependent pathway, but the finding that phorbol esters can also activate Ras in T cells suggested that PKC is involved in an alternative activation pathway (131). However, recent evidence suggests that the immediate target for the Ras-activating effect of phorbol ester is not PKC but, rather, Ras-GRP, a novel phorbol ester–activated Ras guanine nucleotide exchange factor (132). Nevertheless, several studies suggested a link between PKCθ and Ras. First, some PKCθ -mediated functions, e.g., AP-1 activation (36) or upregulation of CD69 expression (M. Villalba, unpublished observation), are blocked by a dominant negative Ras mutant; however, PKCθ-mediated JNK activation was found to be Ras-independent (52). Second, constitutively active PKCθ, just like active Ras (133), can synergize with Ca2+ ionophore or with calcineurin to activate NFAT and the IL-2 promoter (52, 53). Finally, upregulation of CD69 expression on T cells, which is Ras-dependent (134), is deficient in PKCθ knockout T cells (49) or in T cells expressing a dominant-negative PKCθ mutant (122). The most plausible model to account for these findings is that PKCθ acts upstream of Ras, or perhaps in a parallel essential pathway. However, details of the functional connection between these two proteins are presently unknown.

PKCθ -INTERACTING PROTEINS The different PKC isoforms expressed in individual cells are likely to be involved in different biochemical processes and distinct receptor-coupled signaling pathways. Therefore, they require specific mechanisms to ensure that different isoenzymes are targeted to selected locations within the cell, where particular substrates need to be phosphorylated. Intracellular targeting of PKCs is mediated, in part, by specific PKC-binding proteins, which may also affect substrate availability and serve to integrate PKC with other signaling pathways (135, 136). Based on their mode of interaction with PKC and ability to undergo phosphorylation, PKC-binding proteins were divided into three major groups. Receptors for active C kinase (RACK) and receptors for inactive C kinase (RICK) bind active or nonactive enzymes, respectively, but are not targets for phosphorylation by

24 Jan 2002

12:17

776

AR

ISAKOV

AR152-23.tex

¥

AR152-23.SGM

LaTeX2e(2001/05/10)

P1: GJC

ALTMAN

PKC. Although RACK proteins for lipid-dependent PKC isoforms reside within or in close proximity to cell membranes, RICK proteins reside predominantly in the cytoplasm. A third group includes proteins that are phosphorylated by PKC following their interaction with the active enzyme; these proteins are termed substrates that interact with C kinase (STICK) (135, 136). A relatively small number of PKCθ-interacting proteins have been identified thus far; their exact mode of operation with respect to the regulation of PKCθ is not entirely clear.

Annu. Rev. Immunol. 2002.20:761-794. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

14-3-3τ The first PKCθ-binding protein identified is 14-3-3τ , a member of a large family of conserved regulatory proteins expressed in all eukaryotic cells (137, 138). 14-3-3 proteins bind serine-phosphorylated sequences in a multitude of functionally diverse signaling proteins, including kinases, phosphatases, and transmembrane receptors. Their ability to dimerize and simultaneously interact with two or more binding partners enables 14-3-3 proteins to function as scaffolds to facilitate protein-protein interactions and target intracellular protein to specific subcellular locations. The plethora of 14-3-3–interacting proteins allows them to play important roles in a wide range of vital processes, including vesicular transport, cell cycle, mitogenic signal transduction, and programmed cell death. PKCθ was found to coimmunoprecipitate with 14-3-3τ in Jurkat T cell lysates and to interact with immobilized glutathione S–transferase (GST)-14-3-3τ in a pull-down assay or with soluble GST-14-3-3τ in a Far Western overlay assay (37). 14-3-3τ is predominantly cytosolic, and its overexpression in Jurkat cells inhibited phorbol ester–induced cytosol-to-membrane translocation of PKCθ . Overexpression of 14-3-3τ also inhibited PKCθ -dependent IL-2 production, which suggests a RICK-like activity for 14-3-3τ . This activity is further substantiated by findings that a membrane-targeted form of 14-3-3τ increased PKCθ localization in the particulate fraction of resting or stimulated cells (37). It is possible therefore that 14-3-3τ binds PKCθ only in its inactive conformation and thereby targets PKCθ to the cytosol and/or protects it from proteolysis.

Cbl Cbl is a ubiquitously expressed cytoplasmic protein that is abundant in the thymus and cells of the hematopoietic system (139–141). This protein possesses a highly conserved N-terminal phosphotyrosine-binding (PTB) domain, a RING finger motif, a large proline-rich region, a leucine zipper, and a number of potential tyrosine phosphorylation sites. Cbl serves as a substrate of receptor and nonreceptor protein tyrosine kinases and functions as a scaffold protein that forms constitutive and inducible associations with a wide range of signaling intermediates. These include the Grb2 and Crk adapter proteins, the regulatory p85 subunit of PI3K, and the Vav and C3G guanine nucleotide exchange factors for Rac and Rap1 small G-proteins, respectively (140). Furthermore, recent studies demonstrated that Cbl can function as a RING-type, E2-dependent ubiquitin-protein ligase (142).

24 Jan 2002

12:17

AR

AR152-23.tex

AR152-23.SGM

LaTeX2e(2001/05/10)

P1: GJC

Annu. Rev. Immunol. 2002.20:761-794. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

PKCθ IN T CELL ACTIVATION

777

Cbl is a central player in the regulation of protein tyrosine kinase–mediated signaling pathways in TCR-stimulated T cells, where it apparently functions as a negative regulator (140, 141, 143). Analysis of Jurkat T cells demonstrated that Cbl associates weakly with 14-3-3 proteins in unstimulated cells—an effect that was greatly enhanced by TCR ligation and by PKC-activating phorbol esters (144, 145). PMA also inhibited tyrosine phosphorylation of Cbl in anti-CD3–stimulated T cells and consequently the ability of Cbl to associate with SH2 domain–containing signaling molecules (146). The effect of PMA on tyrosine phosphorylation of Cbl was reversed by a PKC inhibitor (GF-109203X), which also restored the activationdependent association of Cbl with PI3K and CrkL. Since most cellular effects of PMA are mediated through PKC activation, Liu et al. tested the relationship between PKC and Cbl in human Jurkat T cells. They found that both PKCθ and PKCα physically associate with, and phosphorylate, Cbl (146). Additional studies revealed that a C-terminal serine-rich motif in Cbl, which is critical for PMAinduced 14-3-3 binding, is the target for phosphorylation by PKCθ and PKCα. These results suggest that PKCθ and PKCα can modulate tyrosine and serine phosphorylation of Cbl in activated T cells and thereby control the association of Cbl with signaling intermediates such as SH2 domain–containing proteins and 143-3 proteins. These effects may play a role in the regulation of specific activation events in stimulated T cells.

Fyn and Lck T cells express two major Src-family PTKs, Fyn and Lck, which are essential for the normal development and function of mature effector T cells (147). Like all Src-family members, they are posttranslationally modified by N-terminal myristoylation and palmitoylation, which facilitates their membrane localization (148). Fyn associates with the TCRε subunits, as well as GPI-linked molecules, such as Thy-1 (149, 150), and functions as a positive regulator of TCR signaling and TCR-induced T cell activation (151–154). In contrast, Lck associates predominantly with the cytoplasmic tails of CD4 or CD8 coreceptors (155, 156), and its absence blocks thymocyte development at the CD4−/CD8− double negative stage (154, 157). In addition, lack of Lck severely compromises TCR signal transduction and T cell activation (158). Based on previous findings that the V1 domain of novel PKCs includes a proteinprotein interaction sequence (159), Ron et al. have used this isolated region as a biochemical bait to potentially identify new PKCθ -binding proteins. Their studies revealed that Fyn was the most prominent tyrosine-phosphorylated protein associated with PKCθ-V1 (160). PKCθ -Fyn interaction was also observed using the yeast two-hybrid system and reciprocal coimmunoprecipitation from T cell lysates. When tested in vitro, PKCθ was found to be a substrate for Fyn. In addition, the presence of Fyn increased PKCθ catalytic activity. An inhibitor of PKCθ binding to Fyn, TER14687, abrogated PKCθ redistribution in CD3-stimulated T cells and decreased cytokine production in a dose-dependent manner. These results suggest

24 Jan 2002

12:17

Annu. Rev. Immunol. 2002.20:761-794. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

778

AR

ISAKOV

AR152-23.tex

¥

AR152-23.SGM

LaTeX2e(2001/05/10)

P1: GJC

ALTMAN

that PKCθ association with Fyn is a physiologically meaningful event in activated T cells, which contributes to upregulation of PKCθ activity and may possibly increase PKCθ translocation to the membrane and triggering of signaling pathways leading to cytokine production. The association between PKCθ and Fyn is potentially interesting because Fyn was recently identified as the tyrosine kinase that plays a major role in phosphorylating and activating Vav under physiological conditions of antigen stimulation (161). This finding and the established functional link between Vav and PKCθ (88, 122) suggest that a tight functional relationship, which is highly relevant to actin cytoskeleton organization and IS assembly, exists among PKCθ , Vav, and Fyn. This notion is also supported by the finding that clustering of transgenic human CD2 molecules on resting mouse T lymphocytes results in Fyn-dependent downstream activation signals, which, in addition to Ca2+ mobilization and MAP kinase activation, also include Vav phosphorylation and PKCθ activation (162). The importance (and nonredundant role) of Fyn in this process is evident from the finding that these responses were decreased in Fyn-deficient T cells, which express normal levels of Lck (162). A second Src family member reported to mediate physical interaction with PKCθ is the Lck kinase. An observation by Liu et al., which led to this finding, revealed that T cell activation is followed by tyrosine phosphorylation of PKCθ, predominantly on Tyr90 in the regulatory domain (117). Phosphorylation was mediated by Lck, which also interacted directly with the PKCθ regulatory domain as demonstrated by pull-down with GST fusion proteins, coimmunoprecipitation, and an overlay assay. Lck association with PKCθ could be observed in resting cells, increased following T cell activation, and involved both the SH2 and SH3 domains of Lck. To further analyze the potential effect of the Lck-mediated tyrosine phosphorylation of PKCθ on T cell activation, Liu et al. replaced Tyr90 with Phe and tested the effects of this mutant on PKCθ -dependent functions. They found that overexpressed constitutively active PKCθ (PKCθ-A148E) increased the proliferation rate of Jurkat cells and synergized with ionomycin in induction of NFAT activity. In contrast, a Tyr90-to-Phe mutation markedly reduced both activities (117). These results suggest that the physical association of Lck with PKCθ and the Lck-induced tyrosine phosphorylation of PKCθ represent physiological events that regulate PKCθ during TCR-induced T cell activation. Therefore PKCθ can physically and functionally interact with both Lck and Fyn, perhaps leading to distinct functional outcomes, depending on the cellular context and the receptor being engaged.

Akt/PKB The Akt/PKB Ser/Thr kinase lies in the crossroads of multiple cellular signaling pathways and acts as a transducer of many functions initiated by receptors that activate PI3K. Akt/PKB is particularly important in mediating cell survival and several metabolic actions of insulin (163, 164). Akt/PKB was recently found to activate the NF-κB signaling pathway in T cells (165, 166) in a manner apparently distinct from that of PKCθ. For example, constitutively active PKCθ activates

24 Jan 2002

12:17

AR

AR152-23.tex

AR152-23.SGM

LaTeX2e(2001/05/10)

P1: GJC

Annu. Rev. Immunol. 2002.20:761-794. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

PKCθ IN T CELL ACTIVATION

779

IKKβ, but not IKKα, in a manner independent of an additional stimulus (79, 80). In contrast, Akt/PKB activates both IKKs and requires additional TCR and/or phorbol ester signals (165, 166). Given the ability of both kinases to activate NFκB and CD28RE, two recent studies addressed the relationship between these two kinases. The first study reported that Akt/PKB mimics the CD28 costimulatory signal leading to NF-κB activation and that it synergizes with PKCθ in this pathway (119). The second study also demonstrated this synergy and, in addition, showed that PKCθ and Akt/PKB constitutively associate in intact T cells and also bind directly to each other in vitro (120). Because both PKCθ and Akt/PKB are recruited to the plasma membrane in activated T cells, their complex is likely to exist in this compartment. The mechanism of cooperation between PKCθ and Akt/PKB in the pathway leading to NF-κB activation in T cells is not fully understood, although the two kinases do not appear to mutually phosphorylate each other (120).

PICOT The human PICOT protein was initially identified by Witte et al. in a study aimed at discovering new PKCθ-binding proteins in activated human T cells. Utilizing the yeast two-hybrid system, they tested binding of a bait consisting of catalytically inactive PKCθ (PKCθ-K409R) to protein products of a Jurkat cell cDNA library (167). The majority of positive clones obtained were found to possess sequences corresponding to a novel gene, which was cloned (GenBank accession no. AAF28844) and further characterized. PICOT is 335 amino acids long; it consists of an amino-terminal thioredoxin (Trx) homology domain, which is required for interaction with PKCθ. PICOT is unlikely to possess classical Trx activity because it lacks one of two critical cysteine residues within the enzyme’s conserved catalytic domain. PICOT is predominantly a cytosolic protein and not a substrate for PKCθ. It may therefore be considered as a putative PKCθ -specific RICK protein. The carboxy terminus of PICOT consists of two tandem repeats of a conserved sequence of 84 residues, termed PICOT-HD (ProDom domain, PD-005573) (168). A position-specific iterative (PSI) Blast search of the NCBI nonredundant database for PICOT-HD–containing sequences revealed more than 40 proteins possessing an expected (E ) value of more than 1e−12, with broad taxonomic distribution, including mammals, insects, yeast, bacteria, and plants. A partial list of these proteins appears in the EMBL database (alignment number: ds44360). Initial functional characterization of PICOT revealed that it inhibits PKCθ induced JNK, but not ERK activation, and downregulates PKCθ-dependent activation of AP-1 and NF-κB in TCR-stimulated Jurkat T cells (167). Because AP-1 and NF-κB are usually activated by various stress signals, these functional effects of PICOT and the conservation of the Trx system and the PICOT-HD domain throughout evolution suggest that PICOT and its relatives may have evolved as proteins that regulate stress-induced signaling pathways in other cell types and organisms via their interaction with kinases.

24 Jan 2002

12:17

780

AR

ISAKOV

AR152-23.tex

¥

AR152-23.SGM

LaTeX2e(2001/05/10)

P1: GJC

ALTMAN

Annu. Rev. Immunol. 2002.20:761-794. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

Nef The HIV Nef protein is expressed early after viral infection and is essential for efficient viral infectivity (169). Promotion of HIV replication by Nef has been explained by its ability to modulate T cell phenotype and function, e.g., downregulate CD4 and MHC class I expression and alter T cell signaling pathways. Analysis of the effects of Nef on different activation-associated events in T cells yielded conflicting results. Some studies demonstrated that Nef inhibits T cell activation at the level of NF-κB, AP-1, and, ultimately, IL-2 induction (170–174), although others reported that Nef promotes T cell activation and increases IL-2 production by activating both NFAT and NF-κB in a manner dependent on its myristoylation and membrane localization (175, 176). However, it appears that enhancement of T cell activation by Nef is physiologically more relevant because thymocytes of Nef-transgenic mice are hyper-responsive to anti-CD3 stimulation (177, 178). Furthermore, the promoting effect of Nef on HIV replication is more consistent with enhanced T cell activation. Early reports of phorbol ester–induced phosphorylation of Nef and its association with Ser/Thr kinase activity (179–182) led Smith et al. to examine potential associations of Nef with PKC enzymes (183). Using coimmunoprecipitation or a GST-Nef fusion protein in pull-down experiments, they found that among five PKC isoenzymes expressed in Jurkat T cells, only PKCθ associated with Nef. However, only ∼5% of the total cytosolic PKCθ interacted with Nef, and the in vitro interaction was enhanced by phospholipids or lipids such as phosphatidylserine and DAG. Indirect support for the possible modulation of PKCθ by Nef was obtained by analysis of PKCθ content and distribution in Nef-expressing Jurkat cells. Nef increased the proportion of PKCθ in the particulate fraction without altering the cellular distribution of other PKC enzymes. In mitogen-stimulated cells, Nef expression resulted in a rapid loss of membrane-residing immunoreactive PKCθ but not PKCβ, -ε, or -ζ (183). Although the actual mechanism of Nef-induced reduction of PKCθ levels in activated T cells is not clear, one possible explanation is that Nef competes with some cellular PKCθ-binding protein and that Nef-bound PKCθ is more susceptible to degradation. These results suggest that Nef-induced changes in the content and distribution of PKCθ could partially account for the functional alterations in HIV-infected T cells. Furthermore, the results indirectly provide additional support for a potential role of PKCθ in T cell activation.

REGULATION OF ACTIVATION-INDUCED T CELL DEATH BY PKCθ Antigen-induced T cell activation is characterized by massive cell proliferation and clonal expansion of antigen-responsive T cells. Clearance of the antigen leads to termination of the immune response, while a balanced T cell homeostasis is regained by a specific mechanism of activation-induced T cell death (AICD) (184, 185).

24 Jan 2002

12:17

AR

AR152-23.tex

AR152-23.SGM

LaTeX2e(2001/05/10)

P1: GJC

Annu. Rev. Immunol. 2002.20:761-794. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

PKCθ IN T CELL ACTIVATION

781

This autologous preprogrammed cell death mechanism is tightly regulated by the TCR-linked signaling pathway. AICD is initiated by a TCR-induced signal leading to expression of the Fas ligand (FasL) on the surface of T cells (186, 187). FasL binding to its corresponding receptor, Fas (CD95), which is constitutively expressed on many cell types (188, 189), triggers a pro-apoptotic signaling pathway leading to protein degradation, DNA fragmentation, and finally, cell death (190). A key regulator of cell death and survival is PI3K, which is linked to multiple cell-surface receptors. It is a cytosolic kinase that is recruited to engaged receptors and activates the Akt/PKB kinase (163, 164). Activated Akt/PKB then phosphorylates and promotes BAD interaction with 14-3-3, which sequesters BAD in the cytoplasm (191–193). The resulting inability of BAD to translocate to the nucleus and heterodimerize with BCL-2 or BCL-XL death antagonists abrogates the death-promoting activity of BAD, which leads to cell survival (194). The mechanisms that regulate FasL expression in activated T cells are not fully understood, but they are likely to involve the transcription factors AP-1, NFAT, and NF-κB, all of which are induced in activated T cells and possess corresponding binding elements in the FasL gene promoter (195–199). The observation that FasL expression (200, 201) and AICD (202) can be induced by a combination of PKC-activating phorbol ester plus Ca2+ ionophore suggests a role for PKC in the regulation of FasL expression. This putative PKC-dependent function has recently been studied by several groups that demonstrated a clear linkage between PKCθ and apoptosis. Thus, PKCθ, but not α, ε, or ζ isoenzymes, selectively activated a FasL promoter-reporter gene and upregulated the mRNA or cell-surface expression of endogenous FasL (203, 204). Enzymatically active PKCθ was essential for upregulating FasL expression, an effect that was augmented by calcineurin (203). PKCθ and calcineurin synergized in promoting a caspase 8–mediated Fas/FasL-dependent AICD, which was specifically blocked by overexpression of CrmA, a caspase 8–specific inhibitor (204). Phorbol esters can protect T cells from Fas-induced apoptosis (205–207), and PKC inhibitors promote anti-Fas-induced apoptosis (208). These findings suggest that under certain conditions, PKC can also mediate antiapoptotic signals. Given the fact that CD28 costimulation can provide a survival signal that protects T cells from AICD (209) and also appears to have an important role in productive activation of PKCθ (79), the possibility exists that PKCθ may also provide a T cell survival signal. This question has been analyzed in two recent publications demonstrating that PKCθ can promote T cell survival, predominantly by phosphorylation and inactivation of BAD (210, 211). These studies demonstrated that PKCθ, together with another novel PKC isoform, PKCε, can promote T cell survival by protecting the cells from Fas-induced apoptosis. The corresponding dominant-negative PKC mutants, as well as pharmacological inhibitors of PKC, such as GF109203X or rottlerin, abolished the effects of phorbol ester and promoted Fas-mediated apoptosis. Both PMA and overexpressed constitutively active PKCθ or PKCε induced BAD phosphorylation at Ser112, an effect reversed by pretreatment of cells with

24 Jan 2002

12:17

Annu. Rev. Immunol. 2002.20:761-794. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

782

AR

ISAKOV

AR152-23.tex

¥

AR152-23.SGM

LaTeX2e(2001/05/10)

P1: GJC

ALTMAN

GF109203X. BAD-induced survival signals are inhibited by p90Rsk (212), and overexpression of a catalytically inactive p90Rsk blocked the protective effect of constitutively active PKC and also inhibited BAD phosphorylation on Ser112 (210). These correlative data suggest that the novel PKCθ and PKCε isoforms can rescue T lymphocytes from AICD by promoting p90Rsk-dependent phosphorylation of BAD. Additional studies by Villalba et al. demonstrated that engagement of Fas by specific antibodies led to a transient activation of PKCθ , which was later followed by caspase 3–dependent cleavage of the enzyme (211). In addition, Fas ligation resulted in proteasome-mediated degradation of PKCθ and inactivation of its catalytic fragment—events that preceded the onset of cell apoptosis. It appears, therefore, that PKCθ can play a dual regulatory role in T cell apoptosis: a promoting role by inducing FasL expression and a protective role by providing a BAD/p90Rsk-dependent survival signal. The T cell differentiation status and the contribution of signals by additional activating/inhibitory receptors may affect the balance between these apparently opposing effects and determine the functional outcome.

CONCLUDING REMARKS Recent studies on PKCθ greatly improved our understanding of the selective function of this particular PKC isoform in T cell activation and established its role as the second critical TCR signal that cooperates with calcineurin to activate the IL-2 gene. Moreover, PKCθ integrates TCR/CD28 costimulatory signals essential for activation of the NF-κB cascade in T cells. The relatively selective expression and essential function of PKCθ in T cell activation and survival suggest that pharmacological or genetic strategies designed to selectively block the function of PKCθ in cells may be therapeutically useful in several potential scenarios. First, because TCR engagement in the absence of CD28 costimulation can lead to T cell anergy (3–6), inhibition of PKCθ may ablate the CD28 costimulatory signal and, therefore, promote anergy. Second, PKCθ inhibition could potentially abolish a T cell survival signal and thus promote the apoptosis of activated self-reactive T cells in autoimmune diseases. Last, because NF-κB activation appears to be necessary for the efficient replication of HIV-1 in T cells (213), similar strategies that interfere with the function of PKCθ may also inhibit viral replication in activated T cells. However, in order to successfully apply these approaches, it would be important to use regimens that temporarily block PKCθ function at a critical time without inducing overt general immunosuppression. Beyond this prospective use of PKCθ as a drug target, there remain several fundamental questions that need to be resolved. First, what is the mechanism that selectively recruits PKCθ to the cSMAC and what protein(s) or lipid(s) mediate this recruitment? Second, what are the immediate physiological substrates of PKCθ and the critical intermediates in the pathway leading from PKCθ to NF-κB and AP-1 activation? Third, what is the precise connection between PKCθ and Ras activation? The recent information reviewed herein provides a solid foundation for

24 Jan 2002

12:17

AR

AR152-23.tex

AR152-23.SGM

LaTeX2e(2001/05/10)

P1: GJC

PKCθ IN T CELL ACTIVATION

783

future studies that will undoubtedly resolve these questions and uncover additional details of the function and regulation of PKCθ in T cells.

Annu. Rev. Immunol. 2002.20:761-794. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

ACKNOWLEDGMENTS We would like to thank our many colleagues who contributed to the work on PKCθ in our laboratories and provided advice. The work in our laboratories was supported in part by grants from the Israel Science Foundation, the Israel Cancer Research Fund, the German-Israeli Foundation for Scientific Research and Development, and the Israel Cancer Association (N. I.), NIH grants CA35299, GM50819, AI49888 (A. A.), and a USA-Israel Binational Science Foundation grant (N. I. and A. A.). This is publication number 444 from the La Jolla Institute for Allergy and Immunology. Visit the Annual Reviews home page at www.annualreviews.org

LITERATURE CITED 1. Grakoui A, Bromley SK, Sumen C, Davis MM, Shaw AS, Allen PM, Dustin ML. 1999. The immunological synapse: a molecular machine controlling T cell activation. Science 285:221–27 2. Monks CR, Freiberg BA, Kupfer H, Sciaky N, Kupfer A. 1998. Three-dimensional segregation of supramolecular activation clusters in T cells. Nature 395: 82–86 3. Lenschow DJ, Walunas TL, Bluestone JA. 1996. CD28/B7 system of T cell costimulation. Annu. Rev. Immunol. 14:233– 58 4. Greenfield EA, Nguyen KA, Kuchroo VK. 1998. CD28/B7 costimulation: a review. Crit. Rev. Immunol. 18:389–418 5. Powell JD, Ragheb JA, Kitagawa-Sakakida S, Schwartz RH. 1998. Molecular regulation of interleukin-2 expression by CD28 co-stimulation and anergy. Immunol. Rev. 165:287–300 6. Chambers CA, Allison JP. 1999. Costimulatory regulation of T cell function. Curr. Opin. Cell Biol. 11:203–10 7. Viola A, Schroeder S, Sakakibara Y, Lanzavecchia A. 1999. T lymphocyte costimulation mediated by reorganiza-

8.

9.

10.

11.

12.

13.

tion of membrane microdomains. Science 283:680–82 Isakov N, Altman A. 1985. Tumor promoters in conjunction with calcium ionophores mimic antigenic stimulation by reactivation of alloantigen-primed murine T lymphocytes. J. Immunol. 135: 3674–80 Kaibuchi K, Takai Y, Nishizuka Y. 1985. Protein kinase C and calcium ion in mitogenic response of macrophage-depleted human peripheral lymphocytes. J. Biol. Chem. 260:1366–69 Truneh A, Albert F, Golstein P, SchmittVerhulst AM. 1985. Early steps of lymphocyte activation bypassed by synergy between calcium ionophores and phorbol ester. Nature 313:318–20 Truneh A, Albert F, Golstein P, SchmittVerhulst AM. 1985. Calcium ionophore plus phorbol ester can substitute for antigen in the induction of cytolytic T lymphocytes from specifically primed precursors. J. Immunol. 135:2262–67 Nishizuka Y. 1995. Protein kinase C and lipid signaling for sustained cellular responses. FASEB J. 9:484–96 Altman A, Mustelin T, Coggeshall KM.

24 Jan 2002

12:17

784

Annu. Rev. Immunol. 2002.20:761-794. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

14.

15.

16.

17.

18.

19.

20.

21.

AR

ISAKOV

AR152-23.tex

¥

AR152-23.SGM

LaTeX2e(2001/05/10)

P1: GJC

ALTMAN

1990. T lymphocyte activation: a biological model of signal transduction. Crit. Rev. Immunol. 10:347–91 Kuo JF, Anersson RG, Wise BC, Mackerlova L, Salomonsson I, Brackett NL, Katoh N, Shoji M, Wrenn RW. 1980. Calcium-dependent protein kinase: widespread occurrence in various tissues and phyla of the animal kingdom and comparison of effects of phospholipid, calmodulin, and trifluoperazine. Proc. Natl. Acad. Sci. USA 77:7039–43 Ogawa Y, Takai Y, Kawahara Y, Kimura S, Nishizuka Y. 1981. A new possible regulatory system for protein phosphorylation in human peripheral lymphocytes. I. Characterization of a calcium-activated, phospholipid-dependent protein kinase. J. Immunol. 127:1369–74 Niedel JE, Kuhn LJ, Vandenbark GR. 1983. Phorbol diester receptor copurifies with protein kinase C. Proc. Natl. Acad. Sci. USA 80:36–40 Kikkawa U, Takai Y, Tanaka Y, Miyake R, Nishizuka Y. 1983. Protein kinase C as a possible receptor protein of tumorpromoting phorbol esters. J. Biol. Chem. 258:11,442–45 Castagna M, Takai Y, Kaibuchi K, Sano K, Kikkawa U, Nishizuka Y. 1982. Direct activation of calcium-activated, phospholipid-dependent protein kinase by tumor-promoting phorbol esters. J. Biol. Chem. 257:7847–51 Touraine JL, Hadden JW, Touraine F, Hadden EM, Estensen R, Good RA. 1977. Phorbol myristate acetate: a mitogen selective for a T-lymphocyte subpopulation. J. Exp. Med. 145:460–65 Abb J, Bayliss GJ, Deinhardt F. 1979. Lymphocyte activation by the tumorpromoting agent 12-O-tetradecanoylphorbol-13-acetate (TPA). J. Immunol. 122:1639–42 Ledbetter JA, June CH, Martin PJ, Spooner CE, Hansen JA, Meie KE. 1986. Valency of CD3 binding and internalization of the CD3 cell-surface complex con-

22.

23.

24.

25.

26.

27.

28.

29.

trol T cell responses to second signals: distinction between effects on protein kinase C, cytoplasmic free calcium, and proliferation. J. Immunol. 136:3945–52 Isakov N, Altman A. 1987. Human T lymphocyte activation by tumor promoters: role of protein kinase C. J. Immunol. 138:3100–7 Kraft AS, Anderson WB. 1983. Phorbol esters increase the amount of Ca2+, phospholipid-dependent protein kinase associated with plasma membrane. Nature 301:621–23 Farrar WL, Ruscetti FW. 1986. Association of protein kinase C activation with IL 2 receptor expression. J. Immunol. 136:1266–73 Ledbetter JA, Gentry LE, June CH, Rabinovitch PS, Purchio AF. 1987. Stimulation of T cells through the CD3/T-cell receptor complex: role of cytoplasmic calcium, protein kinase C translocation, and phosphorylation of pp60c-src in the activation pathway. Mol. Cell. Biol. 7:650–56 Nel AE, Bouic P, Lattanze GR, Stevenson HC, Miller P, Dirienzo W, Stefanini GF, Galbraith RM. 1987. Reaction of T lymphocytes with anti-T3 induces translocation of C-kinase activity to the membrane and specific substrate phosphorylation. J. Immunol. 138:3519–24 Manger B, Weiss A, Imboden J, Laing T, Stobo JD. 1987. The role of protein kinase C in transmembrane signaling by the T cell antigen receptor complex. Effects of stimulation with soluble or immobilized CD3 antibodies. J. Immunol. 139:2755– 60 Baier G, Telford D, Giampa L, Coggeshall KM, Baier Bitterlich G, Isakov N, Altman A. 1993. Molecular cloning and characterization of PKCθ, a novel member of the protein kinase C (PKC) gene family expressed predominantly in hematopoietic cells. J. Biol. Chem. 268:4997– 5004 Osada S, Mizuno K, Saido TC, Suzuki K, Kuroki T, Ohno S. 1992. A new member

24 Jan 2002

12:17

AR

AR152-23.tex

AR152-23.SGM

LaTeX2e(2001/05/10)

P1: GJC

PKCθ IN T CELL ACTIVATION

Annu. Rev. Immunol. 2002.20:761-794. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

30.

31.

32.

33.

34.

35.

36.

37.

of the protein kinase C family, nPKCθ , predominantly expressed in skeletal muscle. Mol. Cell. Biol. 12:3930–38 Chang JD, Xu Y, Raychowdhury MK, Ware JA. 1993. Molecular cloning and expression of a cDNA encoding a novel isoenzyme of protein kinase C (nPKC). A new member of the nPKC family expressed in skeletal muscle, megakaryoblastic cells, and platelets. J. Biol. Chem. 268:14,208–14 Erdel M, Baier-Bitterlich G, Duba C, Isakov N, Altman A, Utermann G, Baier G. 1995. Mapping of the human protein kinase C-θ (PRKCQ) gene locus to the short arm of chromosome 10 (10p15) by FISH. Genomics 25:595–97 Verma RS, Macera MJ, Krishnamurthy M, Abramson J, Kapelner S, Dosik H. 1987. Chromosomal abnormalities in adult T-cell leukemia/lymphoma (ATL). J. Cancer Res. Clin. Oncol. 113:192–96 Monaco G, Pignata C, Rossi E, Mascellaro O, Cocozza S, Ciccimarra F. 1991. DiGeorge anomaly associated with 10p deletion. Am. J. Med. Genet. 39:215– 16 Baier G, Baier-Bitterlich G, Meller N, Coggeshall KM, Giampa L, Telford D, Isakov N, Altman A. 1994. Expression and biochemical characterization of human protein kinase C-θ . Eur. J. Biochem. 225:195–203 Newton A. 1995. Protein kinase C: structure, function, and regulation. J. Biol. Chem. 270:28,495–98 Baier-Bitterlich G, Uberall F, Bauer B, Fresser F, Wachter H, Grunicke H, Utermann G, Altman A, Baier G. 1996. Protein kinase C-θ isoenzyme selective stimulation of the transcription factor complex AP-1 in T lymphocytes. Mol. Cell. Biol. 16:1842–50 Meller N, Liu YC, Collins TL, Bonnefoy-Berard N, Baier G, Isakov N, Altman A. 1996. Direct interaction between protein kinase Cθ (PKCθ ) and 14-3-3τ in T cells: 14-3-3 overexpression results

38.

39.

40.

41.

42.

43.

44.

45. 46.

47.

785

in inhibition of PKCθ translocation and function. Mol. Cell. Biol. 16:5782–91 Donnelly R, Reed MJ, Azhar S, Reaven GM. 1994. Expression of the major isoenzyme of protein kinase-C in skeletal muscle, nPKCθ, varies with muscle type and in response to fructose-induced insulin resistance. Endocrinology 135:2369–74 Mischak H, Goodnight J, Henderson DW, Osada S, Ohno S, Mushinski JF. 1993. Unique expression pattern of protein kinase C-θ: high mRNA levels in normal mouse testes and in T-lymphocytic cells and neoplasms. FEBS Lett. 326:51–55 Hilgenberg L, Miles K. 1995. Developmental regulation of a protein kinase C isoform localized in the neuromuscular junction. J. Cell. Sci. 108:51–61 Erdbrugger W, Keffel J, Knocks M, Otto T, Philipp T, Michel MC. 1997. Protein kinase C isoenzymes in rat and human cardiovascular tissues. Br. J. Pharmacol. 120:177–86 Bauer B, Krumbock N, Ghaffari-Tabrizi N, Kampfer S, Villunger A, Wilda M, Hameister H, Utermann G, Leitges M, Uberall F, Baier G. 2000. T cell expressed PKCθ demonstrates cell-type selective function. Eur. J. Immunol. 30:3645–54 Meller N, Elitzur Y, Isakov N. 1999. Protein kinase C-θ (PKCθ) distribution analysis in hematopoietic cells: proliferating T cells exhibit high proportions of PKCθ in the particulate fraction. Cell. Immunol. 193:185–93 Meller N, Altman A, Isakov N. 1998. New perspectives on PKCθ, a member of the novel subfamily of protein kinase C. Stem Cells 16:178–92 Curran T, Franza BR Jr. 1988. Fos and Jun: the AP-1 connection. Cell 55:395–97 Karin M, Liu Z, Zandi E. 1997. AP-1 function and regulation. Curr. Opin. Cell Biol. 9:240–46 Crabtree GR, Clipstone NA. 1994. Signal transmission between the plasma membrane and nucleus of T lymphocytes. Annu. Rev. Biochem. 63:1045–83

24 Jan 2002

12:17

Annu. Rev. Immunol. 2002.20:761-794. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

786

AR

ISAKOV

AR152-23.tex

¥

AR152-23.SGM

LaTeX2e(2001/05/10)

P1: GJC

ALTMAN

48. Jain J, Loh C, Rau A. 1995. Transcriptional regulation of the IL-2 gene. Curr. Opin. Immunol. 7:333–42 49. Sun Z, Arendt CW, Ellmeier W, Schaeffer EM, Sunshine MJ, Gandhi L, Annes J, Petrzilka D, Kupfer A, Schwartzberg PL, Littman DR. 2000. PKC-θ is required for TCR-induced NF-κB activation in mature but not immature T lymphocytes. Nature 404:402–7 50. Su B, Jacinto E, Hibi M, Kallunki T, Karin M, Ben-Neriah Y. 1994. JNK is involved in signal integration during costimulation of T lymphocytes. Cell 77: 727–36 51. Avraham A, Jung S, Samuels Y, Seger R, Ben-Neriah Y. 1998. Co-stimulationdependent activation of a JNK-kinase in T lymphocytes. Eur. J. Immunol. 28:2320– 30 52. Werlen G, Jacinto E, Xia Y, Karin M. 1998. Calcineurin preferentially synergizes with PKC-θ to activate JNK and IL2 promoter in T lymphocytes. EMBO J. 17:3101–11 53. Ghaffari-Tabrizi N, Bauer B, Villunger A, Baier-Bitterlich G, Altman A, Utermann G, Uberall F, Baier G. 1999. Protein kinase Cθ , a selective upstream regulator of JNK/SAPK and IL-2 promoter activation in Jurkat T cells. Eur. J. Immunol. 29:132–42 54. Dong C, Yang DD, Tournier C, Whitmarsh AJ, Xu J, Davis RJ, Flavell RA. 2000. JNK is required for effector T-cell function but not for T-cell activation. Nature 405:91–94 55. Weiss L, Whitmarsh AJ, Yang DD, Rincon M, Davis RJ, Flavell RA. 2000. Regulation of c-Jun NH(2)-terminal kinase (Jnk) gene expression during T cell activation. J. Exp. Med. 191:139–46 56. Baldwin AS. 1996. The NF-κB and IκB proteins: new discoveries and insights. Annu. Rev. Immunol. 14:649–81 57. Malinin NL, Boldin MP, Kovalenko AV, Wallach D. 1997. MAP3K-related kinase involved in NF-κB induction by

58.

59.

60.

61.

62.

63.

64.

65.

66.

67.

TNF, CD95 and IL-1. Nature 385:540– 44 Ling L, Cao Z, Goeddel DV. 1998. NFκB-inducing kinase activates IKK-α by phosphorylation of Ser-176. Proc. Natl. Acad. Sci. USA 95:2791–97 Ozes ON, Mayo LD, Gustin JA, Pfeffer SR, Pfeffer LM, Donner DB. 1999. NF-κB activation by tumor necrosis factor requires the AKT serine-threonine kinase. Nature 401:82–85 Lee FS, Peters RT, Dang LC, Maniatis T. 1998. MEKK1 activates both IκB kinase α and IκB β. Proc. Natl. Acad. Sci. USA 95:9319–24 Karin M, Ben-Neriah Y. 2000. Phosphorylation meets ubiquitination: the control of NF-κB activity. Annu. Rev. Immunol. 18:621–63 Regnier CH, Song HY, Gao X, Goeddel DV, Cao Z, Rothe M. 1997. Identification and characterization of an IκB kinase. Cell 90:373–83 Woronicz JD, Gao X, Cao Z, Rothe M, Goeddel DV. 1997. IκB kinase-β: NF-κB activation and complex formation with IκB kinase-a and NIK. Science 278:866– 69 Zandi E, Rothwarf DM, Delhase M, Hayakawa M, Karin M. 1997. The IκB kinase complex (IKK) contains two kinase subunits, IKKα and IKKβ, necessary for IκB phosphorylation and NF-κB activation. Cell 91:243–52 Brown K, Gerstberger S, Carlson L, Franzoso G, Siebenlist U. 1995. Control of IκB-α proteolysis by site-specific, signal-induced phosphorylation. Science 267:1485–88 Traenckner EB, Pahl HL, Henkel T, Schmidt KN, Wilk S, Baeuerle PA. 1995. Phosphorylation of human IκBα on serines 32 and 36 controls IκBα proteolysis and NF-κB activation in response to diverse stimuli. EMBO J. 14:2876– 83 Chen Z, Hagler J, Palombella VJ, Melandri F, Scherer D, Ballard D, Maniatis T.

24 Jan 2002

12:17

AR

AR152-23.tex

AR152-23.SGM

LaTeX2e(2001/05/10)

P1: GJC

PKCθ IN T CELL ACTIVATION

68.

Annu. Rev. Immunol. 2002.20:761-794. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

69.

70.

71.

72.

73.

74.

75.

76.

77.

1995. Signal-induced site-specific phosphorylation targets IκBα to the ubiquitinproteasome pathway. Genes Dev. 9:1586– 97 Scherer DC, Brockman JA, Chen Z, Maniatis T, Ballard DW. 1995. Signalinduced degradation of IκBα requires site-specific ubiquitination. Proc. Natl. Acad. Sci. USA 92:11,259–63 May MJ, Ghosh S. 1998. Signal transduction through NF-κB. Immunol. Today 19:80–88 Weiss A, Littman DR. 1994. Signal transduction by lymphocyte antigen receptors. Cell 76:263–74 Ghosh P, Tan TH, Rice NR, Sica A, Young HA. 1993. The interleukin 2 CD28-responsive complex contains at least three members of the NF-κB family: c-Rel, p50, and p65. Proc. Natl. Acad. Sci. USA 90:1696–1700 Lai JH, Horvath G, Subleski J, Bruder J, Ghosh P, Tan TH. 1995. RelA is a potent transcriptional activator of the CD28 response element within the interleukin 2 promoter. Mol. Cell. Biol. 15:4260–71 Maggirwar SB, Harhaj EW, Sun SC. 1997. Regulation of the interleukin-2 CD28-responsive element by NF-ATp and various NF-κB/Rel transcription factors. Mol. Cell. Biol. 17:2605–14 McGuire KL, Iacobelli M. 1997. Involvement of Rel, Fos, and Jun proteins in binding activity to the IL-2 promoter CD28 response element/AP-1 sequence in human T cells. J. Immunol. 159:1319–27 Shapiro VS, Truitt KE, Imboden JB, Weiss A. 1997. CD28 mediates transcriptional upregulation of the interleukin-2 (IL-2) promoter through a composite element containing the CD28RE and NF-IL2B AP-1 sites. Mol. Cell. Biol. 17:4051– 58 Harhaj EW, Sun SC. 1998. IκB kinases serve as a target of CD28 signaling. J. Biol. Chem. 273:25,185–90 Lin X, Cunningham ET Jr, Mu Y, Geleziunas R, Greene WC. 1999. The

78.

79.

80.

81.

82.

83.

84.

85.

787

proto-oncogene Cot kinase participates in CD3/CD28 induction of NF-κB acting through the NF-κB-inducing kinase and IκB kinases. Immunity 10:271–80 Kempiak SJ, Hiura TS, Nel AE. 1999. The Jun kinase cascade is responsible for activating the CD28 response element of the IL-2 promoter: proof of cross-talk with the IκB kinase cascade. J. Immunol. 162:3176–87 Coudronniere N, Villalba M, Englund N, Altman A. 2000. NF-κB activation induced by T cell receptor/CD28 costimulation is mediated by protein kinase C-θ. Proc. Natl. Acad. Sci. USA 97:3394–99 Lin X, O’Mahony A, Mu Y, Geleziunas R, Greene WC. 2000. Protein kinase C-θ participates in NF-κB activation induced by CD3-CD28 costimulation through selective activation of IκB kinase β. Mol. Cell. Biol. 20:2933–40 Khoshnan A, Bae D, Tindell CA, Nel AE. 2000. The physical association of protein kinase Cθ with a lipid raft-associated inhibitor of κB factor kinase (IKK) complex plays a role in the activation of the NF-κB cascade by TCR and CD28. J. Immunol. 165:6933–40 Rao A, Luo C, Hogan PG. 1997. Transcription factors of the NFAT family: regulation and function. Annu. Rev. Immunol. 15:707–47 Manninen A, Huotari P, Hiipakka M, Renkema GH, Saksela K. 2001. Activation of NFAT-dependent gene expression by Nef: conservation among divergent Nef alleles, dependence on SH3 binding and membrane association, and cooperation with protein kinase C-θ. J. Virol. 75:3034–37 Powell JD, Lerner CG, Ewoldt GR, Schwartz RH. 1999. The −180 site of the IL-2 promoter is the target for CREB/ CREM binding in T cell anergy. J. Immunol. 163:6631–39 Solomou EE, Juang YT, Tsokos GC. 2001. Protein kinase c-q participates in the activation of cyclic AMP-responsive

24 Jan 2002

12:17

788

86.

Annu. Rev. Immunol. 2002.20:761-794. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

87.

88.

89.

90.

91.

92.

93.

94.

AR

ISAKOV

AR152-23.tex

¥

AR152-23.SGM

LaTeX2e(2001/05/10)

P1: GJC

ALTMAN

element-binding protein and its subsequent binding to the −180 site of the IL-2 promoter in normal human T lymphocytes. J. Immunol. 166:5665–74 Altman A, Isakov N, Baier G. 2000. Protein kinase Cθ: a new essential superstar on the T-cell stage. Immunol. Today 21:567–73 Szamel M, Appel A, Schwinzer R, Resch K. 1998. Different protein kinase C isoenzymes regulate IL-2 receptor expression or IL-2 synthesis in human lymphocytes stimulated via the TCR. J. Immunol. 160:2207–14 Hehner SP, Li Weber M, Giaisi M, Dr¨oge W, Krammer PH, Schmitz ML. 2000. Vav synergizes with protein kinase Cθ to mediate IL-4 gene expression in response to CD28 costimulation in T cells. J. Immunol. 164:3829–36 Kang S-M, Beverly B, Tran A-C, Brorson K, Schwartz RH, Lenardo MJ. 1992. Transactivation by AP-1 is a molecular target of T cell clonal anergy. Science 257:1134–38 Li W, Whaley CD, Mondino A, Mueller DL. 1996. Blocked signal transduction to the ERK and JNK protein kinases in anergic CD4+ T cells. Science 271:1272–76 Mondino A, Whaley CD, DeSilva DR, Li W, Jenkins MK, Mueller DL. 1996. Defective transcription of the IL-2 gene is associated with impaired expression of c-Fos, FosB, and JunB in anergic T helper 1 cells. J. Immunol. 157:2048–57 Quill H, Riley MP, Cho EA, Casnellie JE, Reed JC, Torigoe T. 1992. Anergic Th1 cells express altered levels of the protein tyrosine kinases p56lck and p59fyn. J. Immunol. 149:2887–93 Gajewski TF, Fields P, Fitch FW. 1995. Induction of the increased Fyn kinase activity in anergic T helper type 1 clones requires calcium and protein synthesis and is sensitive to cyclosporin A. Eur. J. Immunol. 25:1836–42 Boussiotis VA, Barber DL, Lee BJ, Gribben JG, Freeman GJ, Nadler LM.

95.

96.

97.

98.

99.

100.

101.

102.

1996. Differential association of protein tyrosine kinases with the T cell receptor is linked to the induction of anergy and its prevention by B7 family-mediated costimulation. J. Exp. Med. 184:365–76 Salojin K, Zhang J, Cameron M, Gill B, Arreaza G, Ochi A, Delovitch TL. 1997. Impaired plasma membrane targeting of Grb2-murine son of sevenless (mSOS) complex and differential activation of the Fyn-T cell receptor (TCR)-ζ -Cbl pathway mediate the T cell hyporesponsiveness in autoimmune nonobese diabetic mice. J. Exp. Med. 186:887–97 Fields PE, Gajewski TF, Fitch FW. 1996. Blocked Ras activation in anergic CD4+ T cells. Science 271:1276–78 DeSilva DR, Jones EA, Feeser WS, Manos EJ, Scherle PA. 1997. The p38 mitogen-activated protein kinase pathway in activated and anergic Th1 cells. Cell. Immunol. 180:116–23 Boussiotis VA, Freeman GJ, Berezovskaya A, Barber DL, Nadler LM. 1997. Maintenance of human T cell anergy: blocking of IL-2 gene transcription by activated Rap1. Science 278:124–28 Sundstedt A, Sigvardsson M, Leanderson T, Hedlund G, Kalland T, Dohlsten M. 1996. In vivo anergized CD4+ T cells express perturbed AP-1 and NF-κB transcription factors. Proc. Natl. Acad. Sci. USA 93:979–84 Fraser JD, Irving BA, Crabtree GR, Weiss A. 1991. Regulation of interleukin2 gene enhancer activity by the T cell accessory molecule CD28. Science 251: 313–16 Kawasaki H, Nakata Y, Suzuki G, Chihara K, Tokuhisa T, Shiozawa S. 1999. Increased c-Fos/activator protein-1 confers resistance against anergy induction on antigen-specific T cell. Int. Immunol. 11:1873–80 Parra E, McGuire K, Hedlund G, Dohlsten M. 1998. Overexpression of p65 and c-Jun substitutes for B7-1 costimulation by targeting the CD28RE within

24 Jan 2002

12:17

AR

AR152-23.tex

AR152-23.SGM

LaTeX2e(2001/05/10)

P1: GJC

PKCθ IN T CELL ACTIVATION

103. 104.

Annu. Rev. Immunol. 2002.20:761-794. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

105.

106.

107.

108.

109.

110.

111.

112.

113.

114.

115.

the IL-2 promoter. J. Immunol. 160:5374– 81 Paul WE, Seder RA. 1994. Lymphocyte responses and cytokines. Cell 76:241–51 W¨ulfing C, Davis MM. 1998. A receptor/cytoskeletal movement triggered by costimulation during T cell activation. Science 282:2266–69 Penninger JM, Crabtree GR. 1999. The actin cytoskeleton and lymphocyte activation. Cell 96:9–12 Dustin ML, Cooper JA. 2000. The immunological synapse and the actin cytoskeleton: molecular hardware for T cell signaling. Nat. Immunol. 1:23–29 Horejsi V, Drbal K, Cebecauer M, Cerny J, Brdicka T, Angelisova P, Stockinger H. 1999. GPI-microdomains: a role in signalling via immunoreceptors. Immunol. Today 20:356–61 Xavier R, Seed B. 1999. Membrane compartmentation and the response to antigen. Curr. Opin. Immunol. 11:265–69 Janes PW, Ley SC, Magee AI, Kabouridis PS. 2000. The role of lipid rafts in T cell antigen receptor (TCR) signalling. Semin. Immunol. 12:23–34 Bi K, Altman A. 2001. Membrane lipid microdomains and the role of PKCθ in T cell activation. Semin. Immunol. 13:139– 46 Dustin ML, Chan AC. 2000. Signaling takes shape in the immune system. Cell 103:283–94 Monks CR, Kupfer H, Tamir I, Barlow A, Kupfer A. 1997. Selective modulation of protein kinase C-θ during T-cell activation. Nature 385:83–86 Sedwick CE, Blaine K, Miller J. 2001. Costimulation through CD28 is required for targeting of PKCq to the c-SMAC, and activation of NF-κB. Submitted Simons K, Ikonen E. 1997. Functional rafts in cell membranes. Nature 387:569– 72 Moran M, Miceli MC. 1998. Engagement of GPI-linked CD48 contributes to TCR signals and cytoskeletal reorganization: a

116.

117.

118.

119.

120.

121.

122.

123.

124.

789

role for lipid rafts in T cell activation. Immunity 9:787–96 Bi K, Tanaka Y, Coudronniere N, Sugie K, Hong S, van Stipdonk MJ, Altman A. 2001. Antigen-induced translocation of PKC-θ to membrane rafts is required for T cell activation. Nat. Immunol. 2:556–63 Liu Y, Witte S, Liu YC, Doyle M, Elly C, Altman A. 2000. Regulation of protein kinase Cθ function during T cell activation by Lck-mediated tyrosine phosphorylation. J. Biol. Chem. 275:3603–9 Herndon TM, Shan XC, Tsokos GC, Wange RL. 2001. ZAP-70 and SLP-76 regulate protein kinase C-θ and NF-κB activation in response to engagement of CD3 and CD28. J. Immunol. 166:5654– 64 Kane LP, Andres PG, Howland KC, Abbas AK, Weiss A. 2001. Akt provides the CD28 costimulatory signal for upregulation of IL-2 and IFN-γ but not TH2 cytokines. Nat. Immunol. 2:37–44 Bauer B, Krumb¨ock N, Fresser F, Hochholdinger F, Spitaler M, Simm A, ¨ Uberall F, Schraven B, Baier G. 2001. Complex formation and cooperation of PKC and Akt1/PKBa in the NF-κB transactivation cascade in Jurkat T cells. J. Biol. Chem. 276:31,627–34 Dustin ML, Shaw AS. 1999. Costimulation: building an immunological synapse. Science 283:649–50 Villalba M, Coudronniere N, Deckert M, Teixeiro E, Mas P, Altman A. 2000. A novel functional interaction between Vav and PKCθ is required for TCR-induced T cell activation. Immunity 12:151–60 Fischer KD, Kong YY, Nishina H, Tedford K, Marengere LE, Kozieradzki I, Sasaki T, Starr M, Chan G, Gardener S, Nghiem MP, Bouchard D, Barbacid M, Bernstein A, Penninger JM. 1998. Vav is a regulator of cytoskeletal reorganization mediated by the T-cell receptor. Curr. Biol. 8:554–62 Holsinger LJ, Graef IA, Swat W, Chi T, Bautista DM, Davidson L, Lewis RS,

24 Jan 2002

12:17

790

125.

Annu. Rev. Immunol. 2002.20:761-794. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

126.

127.

128.

129.

130.

131.

132.

AR

ISAKOV

AR152-23.tex

¥

AR152-23.SGM

LaTeX2e(2001/05/10)

P1: GJC

ALTMAN

Alt FW, Crabtree GR. 1998. Defects in actin-cap formation in Vav-deficient mice implicate an actin requirement for lymphocyte signal transduction. Curr. Biol. 8:563–72 Pietromonaco SF, Simons PC, Altman A, Elias L. 1998. Protein kinase C-θ phosphorylation of moesin in the actin-binding sequence. J. Biol. Chem. 273:7594–603 Serrador JM, Alonso Lebrero JL, del Pozo MA, Furthmayr H, Schwartz Albiez R, Calvo J, Lozano F, Sanchez Madrid F. 1997. Moesin interacts with the cytoplasmic region of intercellular adhesion molecule-3 and is redistributed to the uropod of T lymphocytes during cell polarization. J. Cell. Biol. 138:1409–23 Rebres RA, Green JM, Reinhold MI, Ticchioni M, Brown EJ. 2001. Membrane raft association of CD47 is necessary for actin polymerization and protein kinase Cθ translocation in its synergistic activation of T cells. J. Biol. Chem. 276:7672– 80 Villalba M, Bi K, Rodriguez F, Tanaka Y, Schoenberger S, Altman A. 2001. Vav1/ Rac-dependent actin cytoskeleton reorganization is required for lipid raft clustering in T cells. J. Cell Biol. 155:331–38 Han J, Luby Phelps K, Das B, Shu X, Xia Y, Mosteller RD, Krishna UM, Falck JR, White MA, Broek D. 1998. Role of substrates and products of PI 3-kinase in regulating activation of Rac-related guanosine triphosphatases by Vav. Science 279:558– 60 Aghazadeh B, Lowry WE, Huang X-Y, Rosen MK. 2000. Structural basis for the relief of autoinhibition of the Dbl homology domain of proto-oncogene Vav by tyrosine phosphorylation. Cell 102:625–33 Downward J, Graves J, Cantrell DA. 1992. The regulation and function of p21ras in T cells. Immunol. Today 13:89– 92 Ebinu JO, Stang SL, Teixeira C, Bottorff DA, Hooton J, Blumberg PM, Barry M, Bleakley RC, Ostergaard HL, Stone JC.

133.

134.

135.

136. 137.

138.

139.

140.

141.

142.

143.

144.

145.

2000. RasGRP links T-cell receptor signaling to Ras. Blood 95:3199–203 Cantrell DA. 1996. T cell antigen receptor signal transduction pathways. Cancer Surv. 27:165–75 D’Ambrosio D, Cantrell DA, Frati L, Santoni A, Testi R. 1994. Involvement of p21ras activation in T cell CD69 expression. Eur. J. Immunol. 24:616–20 Ron D, Kazanietz MG. 1999. New insights into the regulation of protein kinase C and novel phorbol ester receptors. FASEB J. 13:1658–76 Jaken S, Parker PJ. 2000. Protein kinase C binding partners. BioEssays 22:245–54 Fu H, Subramanian R, Masters SC. 2000. 14-3-3 proteins: structure, function, and regulation. Annu. Rev. Pharmacol. Toxicol. 40:617–47 Baldin V. 2000. 14-3-3 proteins and growth control. Prog. Cell. Cycle Res. 4:49– 60 Smit L, Borst J. 1997. The Cbl family of signal transduction molecules. Crit. Rev. Oncog. 8:359–79 Liu YC, Altman A. 1998. Cbl: complex formation and functional implications. Cell. Signal. 10:377–85 Thien CB, Langdon WY. 1998. c-Cbl: a regulator of T cell receptor-mediated signalling. Immunol. Cell. Biol. 76:473–82 Joazeiro CA, Wing SS, Huang H, Leverson JD, Hunter T, Liu YC. 1999. The tyrosine kinase negative regulator c-Cbl as a RING-type, E2-dependent ubiquitinprotein ligase. Science 286:309–12 Lupher MLJ, Rao N, Eck MJ, Band H. 1999. The Cbl protooncoprotein: a negative regulator of immune receptor signal transduction. Immunol. Today 20:375–82 Liu YC, Elly C, Yoshida H, Bonnefoy Berard N, Altman A. 1996. Activationmodulated association of 14-3-3 proteins with Cbl in T cells. J. Biol. Chem. 271: 14,591–95 Liu YC, Liu Y, Elly C, Yoshida H, Lipkowitz S, Altman A. 1997. Serine phosphorylation of Cbl induced by phorbol

24 Jan 2002

12:17

AR

AR152-23.tex

AR152-23.SGM

LaTeX2e(2001/05/10)

P1: GJC

PKCθ IN T CELL ACTIVATION

Annu. Rev. Immunol. 2002.20:761-794. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

146.

147.

148.

149.

150.

151.

152.

153.

154.

155.

ester enhances its association with 14-33 proteins in T cells via a novel serinerich 14-3-3-binding motif. J. Biol. Chem. 272:9979–85 Liu Y, Liu YC, Meller N, Giampa L, Elly C, Doyle M, Altman A. 1999. Protein kinase C activation inhibits tyrosine phosphorylation of Cbl and its recruitment of Src homology 2 domain-containing proteins. J. Immunol. 162:7095–101 Perlmutter RM, Levin SD, Appleby MW, Anderson SJ, Alberola-Ila J. 1993. Regulation of lymphocyte function by protein phosphorylation. Annu. Rev. Immunol. 11: 451–59 Resh M. 1994. Myristylation and palmitylation of Src family members: the fats of the matter. Cell 76:411–13 Samelson LE, Phillips AF, Luong ET, Klausner RD. 1990. Association of the fyn protein-tyrosine kinase with the T-cell antigen receptor. Proc. Natl. Acad. Sci. USA 87:4358–62 Thomas PM, Samelson LE. 1992. The glycophosphatidylinositol-anchored Thy1 molecule interacts with the p60 fyn protein tyrosine kinase in T cells. J. Biol. Chem. 267:12,317–22 Cooke MP, Abraham KM, Forbush KA, Perlmutter RM. 1991. Regulation of T cell receptor signaling by a src family proteintyrosine kinase (p59 fyn). Cell 65:281–91 Appleby MW, Gross JA, Cooke MP, Levin SD, Qian X, Perlmutter RM. 1992. Defective T cell receptor signaling in mice lacking the thymic isoform of p59 fyn. Cell 70:751–63 Stein PL, Lee HM, Rich S, Soriano P. 1992. pp59 fyn mutant mice display differential signaling in thymocytes and peripheral T cells. Cell 70:741–50 Penninger JM, Wallace VA, Kishihara K, Mak TW. 1993. The role of p56lck and p59 fyn tyrosine kinases and CD45 protein tyrosine phosphatase in T-cell development and clonal selection. Immunol. Rev. 135:183–214 Veillette A, Bookman MA, Horak EM,

156.

157.

158.

159.

160.

161.

162.

163.

164.

791

Bolen JB. 1988. The CD4 and CD8 T cell surface antigens are associated with the internal membrane tyrosine-protein kinase p56lck. Cell 55:301–8 Shaw AS, Amrein KE, Hammond C, Stern DF, Sefton BM, Rose JK. 1989. The lck tyrosine protein kinase interacts with the cytoplasmic tail of the CD4 glycoprotein through its unique amino-terminal domain. Cell 59:627–36 Molina TJ, Kishihara K, Siderovski DP, van Ewijk W, Narendran A, Timms E, Wakeham A, Paige CJ, Hartmann KU, Veillette A, Davidson D, Mak TW. 1992. Profound block in thymocyte development in mice lacking p56lck. Nature 357: 161–64 Straus DB, Weiss A. 1992. Genetic evidence for the involvement of the lck tyrosine kinase in signal transduction through the T cell antigen receptor. Cell 70:585– 93 Johnson JA, Gray MO, Chen CH, Mochly-Rosen D. 1996. A protein kinase C translocation inhibitor as an isozymeselective antagonist of cardiac function. J. Biol. Chem. 271:24,962–66 Ron D, Napolitano EW, Voronova A, Vasquez NJ, Roberts DN, Calio BL, Caothien RH, Pettiford SM, Wellik S, Mandac JB, Kauvar LM. 1999. Direct interaction in T-cells between θPKC and the tyrosine kinase p59 fyn. J. Biol. Chem. 274:19,003–10 Huang J, Tilly D, Altman A, Sugie K, Grey H. 2000. T-cell receptor antagonists induce Vav phosphorylation by selective activation of Fyn kinase. Proc. Natl. Acad. Sci. USA 97:10,923–29 Fukai I, Hussey RE, Sunder-Plassmann R, Reinherz EL. 2000. A critical role for p59 fyn in CD2-based signal transduction. Eur. J. Immunol. 30:3507–15 Downward J. 1998. Mechanisms and consequences of activation of protein kinase B/Akt. Curr. Opin. Cell Biol. 10:262–67 Kandel ES, Hay N. 1999. The regulation and activities of the multifunctional

24 Jan 2002

12:17

792

165.

Annu. Rev. Immunol. 2002.20:761-794. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

166.

167.

168.

169.

170.

171.

172.

AR

ISAKOV

AR152-23.tex

¥

AR152-23.SGM

LaTeX2e(2001/05/10)

P1: GJC

ALTMAN

serine/threonine kinase Akt/PKB. Exp. Cell. Res. 253:210–29 Kane LP, Shapiro VS, Stokoe D, Weiss A. 1999. Induction of NF-κB by the Akt/PKB kinase. Curr. Biol. 9:601–4 Jones RG, Parsons M, Bonnard M, Chan VS, Yeh WC, Woodget JR, Ohashi PS. 2000. Protein kinase B regulates T lymphocyte survival, nuclear factor κB activation, and Bcl-X(L) levels in vivo. J. Exp. Med. 191:1721–34 Witte S, Villalba M, Bi K, Liu Y, Isakov N, Altman A. 2000. Inhibition of the cJun N-terminal kinase/AP-1 and NF-κB pathways by PICOT, a novel protein kinase C-interacting protein with a thioredoxin homology domain. J. Biol. Chem. 275:1902–9 Isakov N, Witte S, Altman A. 2000. PICOT-HD: a highly conserved protein domain that is often associated with thioredoxin and glutaredoxin modules. Trends Biochem. Sci. 25:537–39 Geleziunas R, Miller MD, Greene WC. 1996. Unraveling the function of HIV type 1 Nef. AIDS Res. Hum. Retrovir. 12:1579– 82 Greenway AL, McPhee DA, Grgacic E, Hewish D, Lucantoni A, Macreadie I, Azad A. 1994. Nef 27, but not the Nef 25 isoform of human immunodeficiency virus-type 1 pNL4.3 down-regulates surface CD4 and IL-2R expression in peripheral blood mononuclear cells and transformed T cells. Virology 198:245– 56 Luria S, Chambers I, Berg P. 1991. Expression of the type 1 human immunodeficiency virus Nef protein in T cells prevents antigen receptor-mediated induction of interleukin 2 mRNA. Proc. Natl. Acad. Sci. USA 88:5326–30 Bandres JC, Ratner L. 1994. Human immunodeficiency virus type 1 Nef protein down-regulates transcription factors NFκB and AP-1 in human T cells in vitro after T-cell receptor stimulation. J. Virol. 68:3243–49

173. Niederman TM, Garcia JV, Hastings WR, Luria S, Ratner L. 1992. Human immunodeficiency virus type 1 Nef protein inhibits NF-κB induction in human T cells. J. Virol. 66:6213–19 174. Niederman TM, Hastings WR, Luria S, Bandres JC, Ratner L. 1993. HIV-1 Nef protein inhibits the recruitment of AP-1 DNA-binding activity in human T-cells. Virology 194:338–44 175. Schrager JA, Marsh JW. 1999. HIV-1 Nef increases T cell activation in a stimulusdependent manner. Proc. Natl. Acad. Sci. USA 96:8167–72 176. Wang JK, Kiyokawa E, Verdin E, Trono D. 2000. The Nef protein of HIV-1 associates with rafts and primes T cells for activation. Proc. Natl. Acad. Sci. USA 97:394–99 177. Hanna Z, Kay DG, Rebai N, Guimond A, Jothy S, Jolicoeur P. 1998. Nef harbors a major determinant of pathogenicity for an AIDS-like disease induced by HIV-1 in transgenic mice. Cell 95:163–75 178. Skowronski J, Parks D, Mariani R. 1993. Altered T cell activation and development in transgenic mice expressing the HIV-1 nef gene. EMBO J. 12:703–13 179. Sawai ET, Baur A, Struble H, Peterlin BM, Levy JA, Cheng-Mayer C. 1994. Human immunodeficiency virus type 1 Nef associates with a cellular serine kinase in T lymphocytes. Proc. Natl. Acad. Sci. USA 91:1539–43 180. Bodeus M, Marie-Cardine A, Bougeret C, Ramos-Morales F, Benarous R. 1995. In vitro binding and phosphorylation of human immunodeficiency virus type 1 Nef protein by serine/threonine protein kinase. J. Gen. Virol. 76:1337–44 181. Lu X, Wu X, Plemenitas A, Yu H, Sawai ET, Abo A, Peterlin BM. 1996. CDC42 and Rac1 are implicated in the activation of the Nef-associated kinase and replication of HIV-1. Curr. Biol. 6:1677– 84 182. Luo T, Garcia JV. 1996. The association of Nef with a cellular serine/threonine

24 Jan 2002

12:17

AR

AR152-23.tex

AR152-23.SGM

LaTeX2e(2001/05/10)

P1: GJC

PKCθ IN T CELL ACTIVATION

183.

Annu. Rev. Immunol. 2002.20:761-794. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

184.

185. 186.

187.

188.

189.

190. 191.

192.

193.

kinase and its enhancement of infectivity are viral isolate dependent. J. Virol. 70:6493–96 Smith BL, Krushelnycky BW, MochlyRosen D, Berg P. 1996. The HIV Nef protein associates with protein kinase Cθ . J. Biol. Chem. 271:16,753–57 Russell JH. 1995. Activation-induced death of mature T cells in the regulation of immune responses. Curr. Opin. Immunol. 7:382–88 Nagata S. 1997. Apoptosis by death factor. Cell 88:355–65 Dhein J, Walczak H, Baumler C, Debatin KM, Krammer PH. 1995. Autocrine T-cell suicide mediated by APO-1/(Fas/ CD95). Nature 373:438–41 Green DR, Ware CF. 1997. Fas-ligand: privilege and peril. Proc. Natl. Acad. Sci. USA 94:5986–90 Brunner T, Mogil RJ, LaFace D, Yoo NJ, Mahboubi A, Echeverri F, Martin SJ, Force WR, Lynch DH, Ware CF, Green DR. 1995. Cell-autonomous Fas (CD95)/ Fas-ligand interaction mediates activation-induced apoptosis in T-cell hybridomas. Nature 373:441–44 Ju ST, Panka DJ, Cui H, Ettinger R, el Khatib M, Sherr DH, Stanger BZ, Marshak Rothstein A. 1995. Fas(CD95)/FasL interactions required for programmed cell death after T-cell activation. Nature 373:444–48 Nagata S. 1999. Fas ligand-induced apoptosis. Annu. Rev. Genet. 33:29–55 Zha J, Harada H, Yang E, Jockel J, Korsmeyer SJ. 1996. Serine phosphorylation of death agonist BAD in response to survival factor results in binding to 14-3-3 not BCL-XL. Cell 87:619–28 del Peso L, Gonzalez Garcia M, Page C, Herrera R, Nunez G. 1997. Interleukin-3induced phosphorylation of BAD through the protein kinase Akt. Science 278:687– 89 Datta SR, Dudek H, Tao X, Masters S, Fu H, Gotoh Y, Greenberg ME. 1997. Akt phosphorylation of BAD couples survival

194.

195.

196.

197.

198.

199.

200.

201.

202.

793

signals to the cell-intrinsic death machinery. Cell 17:231–41 Yang E, Zha J, Jockel J, Boise LH, Thompson CB, Korsmeyer SJ. 1995. BAD, a heterodimeric partner for Bcl-XL and Bcl-2, displaces Bax and promotes cell death. Cell 80:285–91 Kasibhatla S, Brunner T, Genestier L, Echeverri F, Mahboubi A, Green DR. 1998. DNA damaging agents induce expression of Fas ligand and subsequent apoptosis in T lymphocytes via the activation of NF–κB and AP-1. Mol. Cell. 1:543–51 Latinis KM, Norian LA, Eliason SL, Koretzky GA. 1997. Two NFAT transcription factor binding sites participate in the regulation of CD95 (Fas) ligand expression in activated human T cells. J. Biol. Chem. 272:31,427–34 Holtz Heppelmann CJ, Algeciras A, Badley AD, Paya CV. 1998. Transcriptional regulation of the human FasL promoter-enhancer region. J. Biol. Chem. 273:4416–423 Ivanov VN, Lee RK, Podack ER, Malek TR. 1997. Regulation of Fas-dependent activation-induced T cell apoptosis by cAMP signaling: a potential role for transcription factor NF-κB. Oncogene 14:2455–64 Kasibhatla S, Genestier L, Green DR. 1999. Regulation of fas-ligand expression during activation-induced cell death in T lymphocytes via nuclear factor κB. J. Biol. Chem. 274:987–92 Norian LA, Latinis KM, Koretzky GA. 1998. A newly identified response element in the CD95 ligand promoter contributes to optimal inducibility in activated T lymphocytes. J. Immunol. 161:1078– 82 Latinis KM, Carr LL, Peterson EJ, Norian LA, Eliason SL, Koretzky GA. 1997. Regulation of CD95 (Fas) ligand expression by TCR-mediated signaling events. J. Immunol. 158:4602–11 Rodriguez-Tarduchy G, Sahuquillo AG,

24 Jan 2002

12:17

794

Annu. Rev. Immunol. 2002.20:761-794. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

203.

204.

205.

206.

207.

208.

AR

ISAKOV

AR152-23.tex

¥

AR152-23.SGM

LaTeX2e(2001/05/10)

P1: GJC

ALTMAN

Alarcon B, Bragado R. 1996. Apoptosis but not other activation events is inhibited by a mutation in the transmembrane domain of T cell receptor b that impairs CD3z association. J. Biol. Chem. 271:30,417–25 Villalba M, Kasibhatla S, Genestier L, Mahboubi A, Green DR, Altman A. 1999. Protein kinase cq cooperates with calcineurin to induce Fas ligand expression during activation-induced T cell death. J. Immunol. 163:5813–19 Villunger A, Ghaffari-Tabrizi N, Tinhofer I, Krumbock N, Bauer B, Schneider T, Kasibhatla S, Greil R, Baier-Bitterlich G, Uberall F, Green DR, Baier G. 1999. Synergistic action of protein kinase Cθ and calcineurin is sufficient for Fas ligand expression and induction of a CrmAsensitive apoptosis pathway in Jurkat T cells. Eur. J. Immunol. 29:3549–61 Rodriguez-Tarduchy G, Lopez-Rivas A. 1989. Phorbol esters inhibit apoptosis in IL-2-dependent T lymphocytes. Biochem. Biophys. Res. Commun. 164:1069–75 Kaneko YS, Ikeda K, Nakanishi M. 1999. Phorbol ester inhibits DNA damageinduced apoptosis in U937 cells through activation of protein kinase C. Life Sci. 65:2251–58 G´omez-Angelats M, Bortner DD, Cidlowski JA. 2000. Protein kinase C (PKC) inhibits Fas receptor-induced apoptosis through modulation of the loss of K+ and cell shrinkage. A role for PKC upstream of caspases. J. Biol. Chem. 275:19,609–19 Drew L, Kumar R, Bandyopadhyay D, Gupta S. 1998. Inhibition of the protein

209.

210.

211.

212.

213.

214.

215.

kinase C pathway promotes anti-CD95induced apoptosis in Jurkat T cells. Int. Immunol. 10:877–89 Boise LH, Noel PJ, Thompson CB. 1995. CD28 and apoptosis. Curr. Opin. Immunol. 7:620–25 Bertolotto C, Maulon L, Filippa N, Baier G, Auberger P. 2000. Protein kinase Cθ and ε promote T-cell survival by a Rskdependent phosphorylation and inactivation of BAD. J. Biol. Chem. 275:37,246– 50 Villalba M, Bushway P, Altman A. 2001. Protein kinase C-θ mediates a selective T cell survival signal via phosphorylation of BAD. J. Immunol. 166:5955–63 Tan Y, Ruan H, Demeter MR, Comb MJ. 1999. p90RSK blocks Bad-mediated cell death via a protein kinase C-dependent pathway. J. Biol. Chem. 274:34,859–67 Alcami J, Lain de Lera T, Folgueira L, Pedraza MA, Jacque JM, Bachelerie F, Noriega AR, Hay RT, Harrich D, Gaynor RB. 1995. Absolute dependence on κB responsive elements for initiation and Tatmediated amplification of HIV transcription in blood CD4 T lymphocytes. EMBO J. 14:1552–60 Chow C-W, Rincon M, Cavanagh J, Dickens M, Davis RJ. 1997. Nuclear accumulation of NFAT4 opposed by the JNK signal transduction pathway. Science 278:1638–41 Porter CM, Havens MA, Clipstone NA. 2000. Identification of amino acid residues and protein kinases involved in the regulation of NFATc subcellular localization. J. Biol. Chem. 275:3543–51

Annu. Rev. Immunol. 2002.20:761-794. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

7 Feb 2002 8:47 AR AR152-23-COLOR.tex

See legend on next page.

AR152-23-COLOR.SGM LaTeX2e(2002/01/18) P1: GDL

Annu. Rev. Immunol. 2002.20:761-794. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

7 Feb 2002

8:47

AR

AR152-23-COLOR.tex

AR152-23-COLOR.SGM

LaTeX2e(2002/01/18)

P1: GDL

Figure 1 (See figure on previous page) The central role of PKCθ in TCR/CD28 signaling. Costimulation induces activation of Src-, Syk- and Tec-family PTK, leading to stimulation and membrane recruitment of PLCγ 1, PI3K and Vav (red arrows). PI3K-derived lipids may also be involved in recruiting Vav to the membrane (129). A Vav-mediated pathway, which depends on Rac and actin cytoskeleton reorganization (122) as well as on PI3K (128) is most likely responsible for the selective recruitment of PKCθ to the cSMAC (2), possibly via some undefined, cytoskeleton-associated scaffold. These events appear not to require PLCγ 1-generated DAG (128), although a role for DAG in the initial membrane recruitment of PKCθ cannot be ruled out. PLCγ 1-generated Ca2+ signals also activate calcineurin (CN) via a pathway, which is relatively CD28-independent (green arrow) and is sensitive to immunosuppressive drugs (CsA, FK506). The receptor signals leading to activation of PKCθ and CN can be mimicked by phorbol ester (PMA) and Ca2+ ionophore, respectively (blue arrows). The two major targets of activated PKCθ are NF-κB and AP-1. PKCθ activates IKKβ via unknown intermediates, resulting in stimulation and nuclear translocation of NF-κB. Activation of AP-1 may proceed through SEK1 and JNK, which also require Ca2+/CN signals (50, 51). However, the finding that JNK activation is intact in PKCθdeficient T cells (49) implicates an alternative JNK-independent pathway for AP-1 activation in primary T cells (orange arrow). JNK may even inhibit IL-2 production by primary T cells (54), perhaps via negative regulation of NFAT (214, 215), but the role of PKCθ -dependent JNK activation in IL-2 induction by memory or effector T cells is unresolved. Activated NF-κB and AP-1 bind to their cognate sites in the IL-2 promoter, including the CD28RE, which is a major target of the PKCθ pathway (yellow arrows). In addition, PKCθ can directly phosphorylate and induce the binding of CREB to a cAMP-response element in the IL-2 promoter (85). The functional outcome of this event is unclear, but it may be associated with termination of IL-2 production and/or T cell anergy (5, 84). Yellow rectangles represent active transcription factors, purple lines correspond to inhibitory pathways, and question marks denote potential and/or unresolved pathways. The dashed line indicates the border between the cytoplasm and the nucleus. The structure of the IL-2 promoter is based on Refs. 48, 84.

P1: FRK

February 1, 2002

12:47

Annual Reviews

AR152-FM

Annual Review of Immunology Volume 20, 2002

Annu. Rev. Immunol. 2002.20:761-794. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

CONTENTS FRONTISPIECE, Charles A. Janeway, Jr. A TRIP THROUGH MY LIFE WITH AN IMMUNOLOGICAL THEME, Charles A. Janeway, Jr.

xiv 1

THE B7 FAMILY OF LIGANDS AND ITS RECEPTORS: NEW PATHWAYS FOR COSTIMULATION AND INHIBITION OF IMMUNE RESPONSES, Beatriz M. Carreno and Mary Collins

29

MAP KINASES IN THE IMMUNE RESPONSE, Chen Dong, Roger J. Davis, and Richard A. Flavell

PROSPECTS FOR VACCINE PROTECTION AGAINST HIV-1 INFECTION AND AIDS, Norman L. Letvin, Dan H. Barouch, and David C. Montefiori T CELL RESPONSE IN EXPERIMENTAL AUTOIMMUNE ENCEPHALOMYELITIS (EAE): ROLE OF SELF AND CROSS-REACTIVE ANTIGENS IN SHAPING, TUNING, AND REGULATING THE AUTOPATHOGENIC T CELL REPERTOIRE, Vijay K. Kuchroo, Ana C. Anderson, Hanspeter Waldner, Markus Munder, Estelle Bettelli, and Lindsay B. Nicholson

55 73

101

NEUROENDOCRINE REGULATION OF IMMUNITY, Jeanette I. Webster, Leonardo Tonelli, and Esther M. Sternberg

125

MOLECULAR MECHANISM OF CLASS SWITCH RECOMBINATION: LINKAGE WITH SOMATIC HYPERMUTATION, Tasuku Honjo, Kazuo Kinoshita, and Masamichi Muramatsu

165

INNATE IMMUNE RECOGNITION, Charles A. Janeway, Jr. and Ruslan Medzhitov

KIR: DIVERSE, RAPIDLY EVOLVING RECEPTORS OF INNATE AND ADAPTIVE IMMUNITY, Carlos Vilches and Peter Parham ORIGINS AND FUNCTIONS OF B-1 CELLS WITH NOTES ON THE ROLE OF CD5, Robert Berland and Henry H. Wortis E PROTEIN FUNCTION IN LYMPHOCYTE DEVELOPMENT, Melanie W. Quong, William J. Romanow, and Cornelis Murre

197 217 253 301

LYMPHOCYTE-MEDIATED CYTOTOXICITY, John H. Russell and Timothy J. Ley

SIGNAL TRANSDUCTION MEDIATED BY THE T CELL ANTIGEN x

323

P1: FRK

Annu. Rev. Immunol. 2002.20:761-794. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

February 1, 2002

12:47

Annual Reviews

AR152-FM

CONTENTS

xi

RECEPTOR: THE ROLE OF ADAPTER PROTEINS, Lawrence E. Samelson INTERACTION OF HEAT SHOCK PROTEINS WITH PEPTIDES AND ANTIGEN PRESENTING CELLS: CHAPERONING OF THE INNATE AND ADAPTIVE IMMUNE RESPONSES, Pramod Srivastava CHROMATIN STRUCTURE AND GENE REGULATION IN THE IMMUNE SYSTEM, Stephen T. Smale and Amanda G. Fisher PRODUCING NATURE’S GENE-CHIPS: THE GENERATION OF PEPTIDES FOR DISPLAY BY MHC CLASS I MOLECULES, Nilabh Shastri, Susan

371

Schwab, and Thomas Serwold

395 427

463

THE IMMUNOLOGY OF MUCOSAL MODELS OF INFLAMMATION, Warren Strober, Ivan J. Fuss, and Richard S. Blumberg T CELL MEMORY, Jonathan Sprent and Charles D. Surh

GENETIC DISSECTION OF IMMUNITY TO MYCOBACTERIA: THE HUMAN MODEL, Jean-Laurent Casanova and Laurent Abel ANTIGEN PRESENTATION AND T CELL STIMULATION BY DENDRITIC CELLS, Pierre Guermonprez, Jenny Valladeau, Laurence Zitvogel, Clotilde Th´ery, and Sebastian Amigorena

495 551 581

621

NEGATIVE REGULATION OF IMMUNORECEPTOR SIGNALING, Andr´e Veillette, Sylvain Latour, and Dominique Davidson

669

CPG MOTIFS IN BACTERIAL DNA AND THEIR IMMUNE EFFECTS, Arthur M. Krieg

PROTEIN KINASE Cθ

709 IN

T CELL ACTIVATION, Noah Isakov and Amnon

Altman

RANK-L AND RANK: T CELLS, BONE LOSS, AND MAMMALIAN EVOLUTION, Lars E. Theill, William J. Boyle, and Josef M. Penninger PHAGOCYTOSIS OF MICROBES: COMPLEXITY IN ACTION, David M. Underhill and Adrian Ozinsky

761 795 825

STRUCTURE AND FUNCTION OF NATURAL KILLER CELL RECEPTORS: MULTIPLE MOLECULAR SOLUTIONS TO SELF, NONSELF DISCRIMINATION, Kannan Natarajan, Nazzareno Dimasi, Jian Wang, Roy A. Mariuzza, and David H. Margulies

853

INDEXES Subject Index Cumulative Index of Contributing Authors, Volumes 1–20 Cumulative Index of Chapter Titles, Volumes 1–20

ERRATA An online log of corrections to Annual Review of Immunology chapters (if any, 1997 to the present) may be found at http://immunol.annualreviews.org/

887 915 925

8 Feb 2002

10:11

AR

AR152-24.tex

AR152-24.SGM

LaTeX2e(2001/05/10) P1: IKH 10.1146/annurev.immunol.20.100301.064753

Annu. Rev. Immunol. 2002. 20:795–823 DOI: 10.1146/annurev.immunol.20.100301.064753 c 2002 by Annual Reviews. All rights reserved Copyright °

RANK-L AND RANK: T Cells, Bone Loss, and Mammalian Evolution Lars E. Theill1, William J. Boyle2, and Josef M. Penninger3∗ Annu. Rev. Immunol. 2002.20:795-823. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

1

Inflammation Drug Discovery Research, 2Discovery Research Amgen Inc., One Amgen Center Drive, Thousand Oaks, California 91320-1789; e-mail: [email protected] 3 The Amgen Institute, Ontario Cancer Institute, and the Departments of Medical Biophysics and Immunology, University of Toronto, 620 University Avenue, Toronto, Ontario M5G 2C1, Canada

Key Words T cell–dendritic cell interaction, osteoimmunology, TNF/TNFR super family molecules, osteoclast ■ Abstract TNF and TNFR family proteins play important roles in the control of cell death, proliferation, autoimmunity, the function of immune cells, or the organogenesis of lymphoid organs. Recently, novel members of this large family have been identified that have critical functions in immunity and that couple lymphoid cells with other organ systems such as bone morphogenesis and mammary gland formation in pregnancy. The TNF-family molecule RANK-L (RANK-L, TRANCE, ODF) and its receptor RANK are key regulators of bone remodeling, and they are essential for the development and activation of osteoclasts. Intriguingly, RANK-L/RANK interactions also regulate T cell/dendritic cell communications, dendritic cell survival, and lymph node formation; T cell–derived RANK-L can mediate bone loss in arthritis and periodontal disease. Moreover, RANK-L and RANK are expressed in mammary gland epithelial cells, and they control the development of a lactating mammary gland during pregnancy and the propagation of mammalian species. Modulation of these systems provides us with a unique opportunity to design novel therapeutics to inhibit bone loss in arthritis, periodontal disease, and osteoporosis.

INTRODUCTION More than three decades ago, lymphotoxin and tumor necrosis factor were identified as products of lymphocytes and macrophages that caused the lysis of certain types of cells, especially tumor cells (1–3). Large-scale sequencing efforts allowed for the identification of many related proteins, collectively referred to as TNF- and TNFR-related superfamily proteins. The receptors and ligands in this superfamily have unique structural attributes that couple them directly to signaling pathways for cell proliferation, survival, and differentiation (4). ∗

To whom correspondence should be addressed.

0732-0582/02/0407-0795$14.00

795

4 Feb 2002

12:2

Annu. Rev. Immunol. 2002.20:795-823. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

796

AR

AR152-24.tex

THEILL

¥

BOYLE

AR152-24.SGM

¥

LaTeX2e(2001/05/10)

P1: IKH

PENNINGER

TNF family members are type 2 (carboxy-terminus outside the cell) transmembrane proteins that assemble into functional trimers. The trimeric ligands, either membrane bound or proteolytically cleaved to a soluble form, bind their specific receptors and elicit either growth or cell death. TNF receptors (TNFR) are type 1 transmembrane glycoproteins with a characteristic cysteine-rich extracellular domain. Two subgroups of TNFR homologues, Fas, TNFR1, DR3, DR4, DR5, and DR6, contain intracellular death domain that bind TRADD or FADD. This leads to activation of caspase 8 and apoptosis (4). However, signaling through deathreceptors can also be required for proliferation of hepatocytes and T cells (5, 6). The other group including TNFR2, CD40, or CD30 bind TNF-receptor associated factors (TRAFs), molecular adapters that couple these surface receptors to downstream signaling cascades. This leads to activation of JNK/SAPK and NFκB, which can promote cell growth and survival. These proteins therefore play critical roles in morphogenesis, the control of apoptosis, differentiation, or proliferation. TNF/TNFR superfamily proteins are now extensively studied as targets for therapies against many human diseases such as atherosclerosis, allograft rejection, arthritis, and cancer. For excellent recent reviews see References (4, 7). Members of the tumor necrosis factor superfamilies of ligands and cell-surface receptors regulate immune function, and most TNF/TNFR superfamily proteins, such as FASL/FAS, CD40L/CD40, TNF/TNFR, or LTβ/LTβR, to name a few, are expressed in the immune system, where they coordinate immune cell homeostasis, activation-induced cell death, T cell priming, functions, and survival of dendritic cells, or the formation of germinal centers and lymphoid organs such as Peyer’s patches and lymph nodes (4, 8, 9). Recently, novel members of this large family have been identified that have critical functions in immunity and couple lymphoid cells with other organ systems such as bone morphogenesis and mammary gland formation in pregnancy. In this review we discuss the biological functions of RANK-L, its receptor RANK (receptor activator of NFκB ligand), and the decoy receptor OPG. These proteins have become principal targets in development of new therapies for osteoporosis, tooth loss, arthritis, or bone metastases in cancer.

RANK-L, RANK, OPG: Molecular Links Between Bone Remodeling, Immunity, and Pregnancy The TNF-family molecule RANK-L [ligand to receptor activator of NFκB ligand; also known as osteoprotegerin-ligand (OPG-L); TNF-related activation-induced cytokine (TRANCE), osteoclast differentiation factor (ODF), and TNFSF11] and its receptor RANK (TNFRSF11A) are key regulators of bone remodeling and essential for the development and activation of osteoclasts. RANK-L also regulates T cell/dendritic cell communications, dendritic cell survival, and lymph node organogenesis. Production of RANK-L by activated T cells can directly regulate osteoclastogenesis and bone remodeling, and it explains why autoimmune diseases, cancers, leukemias, asthma, chronic viral infections, and periodontal disease result in systemic and local bone loss. In particular, RANK-L appears to be the pathogenetic principle that causes bone and cartilage destruction in arthritis.

8 Feb 2002

10:44

AR

AR152-24.tex

AR152-24.SGM

LaTeX2e(2001/05/10)

P1: IKH

RANK-L AND RANK

797

Annu. Rev. Immunol. 2002.20:795-823. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

Inhibition of RANK-L function via the natural decoy receptor osteoprotegerin (OPG, TNFRSF11B) prevents bone loss in postmenopausal osteoporosis and cancer metastases and completely blocks crippling in a rat model of arthritis. Intriguingly, RANK-L and RANK play essential roles in the formation of a lactating mammary gland in pregnancy and lactating. Thus, this system provided an unexpected molecular paradigm that links bone morphogenesis, T cell activation, and the organization of lymphoid tissues, with mammary gland formation required for the survival of mammalian species.

Remodeling of Bone Morphogenesis and remodeling of bone involve the synthesis of bone matrix by osteoblasts and the coordinate resorption of bone by osteoclasts (10). In fact, it has been estimated that ∼10% of the total bone mass in humans is being remodeled per year. Osteoblasts and osteoclasts arise from distinct cell lineages and maturation processes, i.e., osteoclasts arise from mesenchymal stem cells, whereas osteoclasts differentiate from hematopoietic monocyte/macrophage precursors (Figure 1) (11). Imbalances between osteoclast and osteoblast activities can arise from a wide variety of hormonal changes or perturbations of inflammatory and growth factors, resulting in skeletal abnormalities characterized by decreased (osteoporosis) or increased (osteopetrosis) bone mass. Increased osteoclast activity is seen in many osteopenic disorders, including postmenopausal osteoporosis, Paget’s disease, lytic bone metastases or rheumatoid arthritis, leading to increased bone resorption and crippling bone damage (10). Various factors have been described including CSF-1 (M-CSF), IL-1, TGF-β, TGF-α, TNFα, TNFβ, IL-6, vitamin 1,25-hihydroxyvitamin D3, IL-11, calcitonin, PGE2, or parathyroid hormone (PTH), all of which affect osteoclastogenesis at distinct stages of development (11). However, genetic ablation experiments have shown that these factors are not essential for osteoclast development in vivo. Due to the enormous social and economic impacts of bone loss, the crippling effects to human health, as well as the search to increase the human life-span without the side-effects of old age, it was of paramount importance to identify essential factors involved in osteoclast development and bone remodeling. These factors are the TNF-TNFR superfamily proteins RANK-L and RANK. We first discuss the genetic and functional links between bone metabolism and the immune system that explain bone loss associated with multiple diseases. Inhibition of RANK-L function via its natural decoy receptor osteoprotegerin (OPG) or small molecules might be the future treatment of choice to abolish osteoporosis, tooth loss, or crippling in arthritis. RANK-L—IDENTIFICATION OF THE CRITICAL OSTEOCLAST DIFFERENTIATION FACTOR

RANK-L/RANKLTRANCE/ODF was cloned simultaneously by four independent groups (12–15). The rankl gene encodes a TNF superfamily molecule of 316 amino acids (38 kDa), and three RANK-L subunits assemble to form the functional trimeric molecule. Trimeric RANK-L is initially made as a membrane

4 Feb 2002

12:2

Annu. Rev. Immunol. 2002.20:795-823. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

798

AR

AR152-24.tex

THEILL

¥

BOYLE

AR152-24.SGM

¥

LaTeX2e(2001/05/10)

P1: IKH

PENNINGER

anchored molecule and can be subsequently released from the cell surface as soluble homotrimeric molecules following proteolytic cleavage by the metalloproteasedisintegrin TNFα convertase (TACE) (16). It remains to be seen whether TACE is indeed the critical protease required for the release of RANK-L from the cell surface. Although slight functional differences may exist, in our hands both soluble and membrane-bound RANK-L can function as potent agonistic ligands for osteoclastogenesis in vitro (13, 16, 17). However, other authors have suggested that membrane-bound RANK-L may work more efficiently than soluble RANK-L (18). RANK-L is extensively expressed in osteoblast/stromal cells, primitive mesenchymal cells surrounding the cartilaginous anlagen and hypertrophied-chondrocytes. RANK-L expression can be upregulated by bone resorbing factors such as glucocorticoids, vitamin D3, IL-1, IL-6, IL-11, IL-17, TNFα, PGE2, or PTH (Table 1) (13, 15). Using in vitro culture systems, it has been shown that RANK-L can both activate mature osteoclasts and mediate osteoclastogenesis in the presence of CSF-1 (13, 15). rankl−/− mice display severe osteopetrosis, stunted growth, and a defect in tooth eruption, and rankl−/− osteoblasts cannot support osteoclastogenesis. However, these mice contain hematopoietic precursors that can differentiate into phenotypically and functionally mature osteoclasts in vitro in the presence of recombinant RANK-L and CSF-1. Importantly, osteoblast cell lines derived from rankl−/− mice do not support osteoclast formation, indicating that the defect

TABLE 1 Molecules that regulate RANK-L and OPG levels RANKL

OPG

Hormones Vitamin-D3 PTH PTHrP Estradiol

Increased Increased Increased No change

Increased Decreased Decreased Increased

Cytokines TNF-α IL-1 IL-6 IL-11 IL-17

Increased Increased Increased Increased Increased

Increased Increased n.t. n.t. n.t.

Growth factors TGF-β BMP-2

Decreased n.t.

Increased Increased

Others Prostaglandin E2 Glucocorticoid CD40L

Increased Increased Increased

Decreased Decreased n.t.

8 Feb 2002

10:47

AR

AR152-24.tex

AR152-24.SGM

LaTeX2e(2001/05/10)

P1: IKH

Annu. Rev. Immunol. 2002.20:795-823. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

RANK-L AND RANK

799

in osteoclastogenesis observed in rankl−/− mice is due to an intrinsic defect in osteoblastic stroma. Whereas csf-1 mutant op/op mice display a developmental arrest in both monocyte/macrophage and osteoclast lineages, rankl−/− mice display normal monocyte/macrophage differentiation and normal differentiation of dendritic cells (DCs). The osteoclast defect in csf-1 mutant op/op mice is not absolute and older op/op mice do have, albeit only few, osteoclasts. Moreover, the defect in op/op mice can be reversed by transgenic overexpression of Bcl-2 in the osteoclast/monocyte lineage indicating that—in contrast to RANK-L—CSF-1 expression is not essential for osteoclast development (19). Thus, RANK-L is a specific and essential differentiation factor for osteoclast precursors and an activation factor for mature osteoclasts (Figure 2).

RANK The receptor for RANK-L is RANK (receptor activator of NFκB ligand, also known as TRANCE-R, or TNFRSF11A), a member of TNF-R superfamily. RANK is expressed as a transmembrane heterotrimer on the surface of hematopoietic osteoclast progenitors, mature osteoclasts, chondrocytes, and mammary gland epithelial cells (12, 20). In vitro ligation of RANK with RANK-L results in osteoclastogenesis from progenitor cells and the activation of mature osteoclasts (20–22). Mice with a genetic mutation of RANK are exact phenocopies of rankl−/− mice and have a complete block in osteoclast development that can be restored by reintroduction of RANK into bone marrow progenitor cells (23, 24). The osteopetrosis observed in these mice can be reversed by transplantation of bone marrow from rag1−/− mice, indicating that rank−/− mice have an intrinsic defect in osteoclast function (24). Thus, the interaction between RANK-L expressed by stromal cells/osteoblasts and its receptor RANK expressed on osteoclast precursors is essential for osteoclastogenesis (Figure 2). In human familial expansile osteolysis, a rare autosomal dominant bone disorder characterized by focal areas of increased bone remodeling (25), a heterozygous insertion mutation in exon 1 of RANK has been noted that appears to increase RANK-mediated NFκB activation and thus might be causal for the disease. These genetic results in humans and mutant mice established the absolute dependency of osteoclast differentiation and activation of mature osteoclasts on the expression of RANK-L and RANK. When RANK on osteoclasts is activated, it sends signals into the cells through adapter proteins (Figure 3). RANK contains 383 amino acids in its intracellular domain (residues 234–616), which contains three putative binding domains (termed I, II, and III) for tumor necrosis factor receptor-associated factors (TRAFs) (26). Indeed, RANK interacts with TRAFs 1, 2, 3, 5, and 6 both in vitro and in cells (27). Mapping of the structural requirements for TRAF/RANK interaction revealed multiple TRAF binding sites clustered in two distinct domains in the RANK cytoplasmic tail. These TRAF binding domains were shown to be functionally important for the RANK-dependent induction of NF-κB and

4 Feb 2002

12:2

Annu. Rev. Immunol. 2002.20:795-823. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

800

AR

AR152-24.tex

THEILL

¥

BOYLE

AR152-24.SGM

¥

LaTeX2e(2001/05/10)

P1: IKH

PENNINGER

c-Jun NH2-terminal kinase (JNK) activities. In particular, TRAF6 interacts with membrane-proximal determinants distinct from those binding TRAFs 1, 2, 3, and 5. When this membrane-proximal TRAF6 interaction domain was deleted, RANKmediated NF-κB signaling was completely inhibited, whereas JNK activation was only partially inhibited (26–29), suggesting that interaction with TRAFs is necessary for NF-κB activation but not essential for activation of the JNK pathway. Indeed mice lacking TRAF6 have bone phenotypes similar to that of rankl−/− and rank−/− mice due to a partial block in osteoclastogenesis and defective activation of mature osteoclasts (30–32). It should be noted that TRAF6 mutant mice still have TRAP+ osteoclasts (31), whereas NF-κB1/NF-κB2 double mutant mice lack TRAP+ osteoclasts (33, 34). Thus, TRAF6 is a critical factor involved in the activation of mature osteoclasts, but other TRAF6s (and possibly other molecules) appear to be able to partially substitute for the loss of TRAF6 during osteoclast development. In line with these data, osteoclastogenesis can be initiated in rank−/− mice by transfer of mutant RANK that lacks the TRAF6 binding site (24, and unpublished). RANK-L also activates the anti-apoptotic serine/threonine kinase Akt/PKB through a signaling complex involving c-Src and TRAF6 (35). c-Src and TRAF6 interact with each other and with RANK following receptor engagement, and a deficiency in c-Src or the addition of Src family kinase inhibitors blocks TRANCEmediated Akt/PKB activation in osteoclasts. TRAF6, in turn, enhances the kinase activity of c-Src leading to tyrosine phosphorylation of downstream signaling molecules such as c-Cbl (35). Moreover, RANK can recruit TRAF6, Cbl family scaffolding proteins, and the phospholipid kinase PI3-K in a ligand- and Srcdependent manner. RANK-L mediated Akt/PKB activation is defective in cbl-b−/− dendritic cells (36). These findings implicate Cbl family proteins not only as negative regulators of signaling, but also as positive modulators of TNFR superfamily signaling. Moreover, these data provided the first evidence of a cross-talk between TRAF proteins and Src family kinases. In addition, it should be noted that inhibition of p38 kinases using SB203580 and overexpression of dominant negative p38α or MKK6 inhibit RANK-L-induced differentiation of the osteoclast-like cell line RAW264 (37).

Osteoprotegerin—Protector of the Bone Osteoprotegerin (OPG, “protector of the bone”; also known as osteoclastogenesis inhibitory factor, OCIF) is a secreted protein with homology to members of the TNF receptor family (13, 15, 38, 39). The opg gene encodes a 44-kDa protein that is posttranslationally modified to a 55-kDa molecule through N-linked glycosylation. Although OPG is a member of the TNFR-family, whose membrane normally assembles as molecular trimers, OPG is secreted as a 110-kDa homodimer. OPG functions as a soluble decoy receptor to RANK-L and competes with RANK for RANK-L binding. Consequently, OPG is an effective inhibitor of osteoclast maturation and activation in vitro (13, 38). High systemic levels of osteoprotegerin

8 Feb 2002

10:48

AR

AR152-24.tex

AR152-24.SGM

LaTeX2e(2001/05/10)

P1: IKH

Annu. Rev. Immunol. 2002.20:795-823. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

RANK-L AND RANK

801

(OPG) in OPG transgenic mice cause osteopetrosis with normal tooth eruption and bone elongation, and these levels also inhibit the development and activity of endosteal, but not periosteal, osteoclasts (38). By contrast, OPG-deficient mice display severe osteoporosis associated with a high incidence of fractures (40, 41), indicating that the level of bone mass correlates with the levels of OPG in mice. Expression of OPG in ST2 stromal cell line and human bone marrow stromal cells is downregulated by bone-resorbing factors such as vitamin D3 [1,25(OH)2D3], prostaglandin E2 (PGE2), or glucocorticoids and is upregulated by Ca2+ ions and TGFβ (42, 43, 39). OPG is also expressed in follicular dendritic cells (identified as FDCR-1) and is upregulated following CD40 ligating (44). In addition to osteoporosis, some but not all OPG mutant mice develop calcification of their large arteries (40), and RANK-L and RANK transcripts are detected in the calcified arteries of OPG−/− mice (45). Transgenic OPG delivered from midgestation through adulthood does prevent the formation of arterial calcification in opg−/− mice by blocking a process resembling osteoclastogenesis (45). These data indicate that the OPG/RANK-L/RANK signaling pathway may play an important role in both pathological and physiological calcification processes. Such findings may also explain the observed high clinical incidence of vascular calcification in the osteoporotic patient population (46). Since woman with osteoporosis have increased incidence of strokes (47, 48), OPG, RANK-L, and RANK may play, similar to the CD40 and CD40L system (49), a role in the pathogenesis of atherosclerosis, strokes, or heart attacks via a yet unknown regulation of endothelial cells. All genetic and functional experiments by many different groups indicate that the balance between RANK-L-RANK signaling and the levels of biologically active OPG regulate development and activation of osteoclasts and bone metabolism (Figure 2). Intriguingly, all factors that inhibit or enhance bone resorption via osteoclasts act via regulation of RANK-L-RANK and/or OPG. Thus, it appears that the complex system of osteoclast-regulated bone-remodeling is only controlled by these three molecule. However, although RANK-L is also expressed in many other tissues than the bone, osteoclast development is restricted to the bone microenvironment suggesting that another tissue-specific factor may exist that acts in concert with RANK-L/RANK. It has recently been shown in vitro that TNFα and IL-1 can apparently induce the development of TRAP+ osteoclasts in the absence of RANK/RANK-L (50, 51). However, in our own genetic experiments using RANK-deficient osteoclast progenitors, TNFα as well as IL-1-dependent osteoclastogenesis are strictly dependent on RANK expression. Thus, whereas TNFα and IL-1 appear to potentiate the development of osteoclasts (52), presumably via activation of common second messenger systems such as NFκB activation, both of these molecules rely on the expression of RANK-L/RANK. It should be also noted that mutations of TNFα, TNFR1, or TNFR2 do not cause any alterations in bone metabolisms or osteoclast development/activation in vivo. In addition to the association between RANK-L and OPG, OPG can also bind to the TNF-family molecule TRAIL at low stoichiometry (∼10,000 times less binding to TRAIL than to RANK-L) (53). OPG-Fc binds TRAIL with an affinity of 3.0 nM,

8 Feb 2002

10:49

Annu. Rev. Immunol. 2002.20:795-823. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

802

AR

THEILL

AR152-24.tex

¥

BOYLE

¥

AR152-24.SGM

LaTeX2e(2001/05/10)

P1: IKH

PENNINGER

which is slightly weaker than the interaction of TRID-Fc or DR5- Fc with TRAIL. Functionally, high doses of OPG inhibit TRAIL-induced apoptosis of Jurkat cells and TRAIL can block the anti-osteoclastogenic activity of OPG (53). These data suggest potential cross-regulatory mechanisms by OPG and TRAIL. However, it is still not known whether OPG-TRAIL interactions have any functional relevance in vivo. Importantly, OPG expression is induced by estrogen in cell lines and in vivo (54, 55), which might explain postmenopausal osteoporosis in women: That is, reduced ovarian function leads to reduced estrogen levels and hence reduced OPG levels, which release RANK-L from the inhibition by the decoy receptor. Injection of OPG into ovariectomized female rats blocks bone loss and osteoporosis normally associated with the loss of ovarian function (38). Thus, OPG and/or modulation of RANK-L-RANK function via small molecules are promising avenues to prevent postmenopausal osteoporosis. In essence, OPG appears to function in bone loss similar to insulin in diabetes: Injection of OPG prevents osteoclast activation and osteopenia in essentially every model system of osteoclast-mediated bone loss.

The Role of RANK-L and RANK in the Immune System At around the same time we had the first evidence that RANK-L might play a role in osteoclast development, RANK-L (TRANCE) was independently cloned by two other groups as a molecule expressed on the surface of activated T cells (12, 14). Both soluble and membrane-bound RANK-L is produced by activated CD4+ and CD8+ T cells (14, 17). RANK-L is also expressed in lymph nodes, spleen, thymus and intestinal lymphoid patches (13), and immature CD4−CD8− thymocytes (12). RANK-L expression in T cells is induced by antigen receptor engagement and is regulated by calcineurin, ERK1/ERK-2, and PKC-regulated signaling pathways (14, 17). RANK is expressed on the surface of dendritic cells (DCs), mature T cells, and hematopoietic precursors; RANK-L-RANK interactions can induce cluster formation, Bcl-XL expression, survival, CD40 expression, and IL-12 production in DC (Figure 4) (12, 14, 56). In addition, OPG was found in a screen to identify novel genes expressed in follicular dendritic cells. OPG can be found on the cell surface of DCs probably by capturing of soluble OPG the cell membrane via binding of a hyalurinic acid binding region present in OPG (44). Thus, like the interactions between CD40-L and CD40, or CD28 and CD80/CD86, the binding of RANK-L to RANK can regulate DC functions, T cell activation, and T cell-DC communication in vitro (12, 57). Moreover, OPG may modulate this interaction.

Lymph Node Organogenesis During the initial analyses of rankl−/− and later of rank mutant mice, a completely unexpected phenotype became evident: rankl−/− and rank−/− mice displayed complete absence of all lymph nodes (23, 24, 58). Recent studies of mice deficient for lymphotoxin-α (LTα ) (59, 60), LTβ (61, 62), TNF-R1 (TNFRp55) (63), LTβ receptor (LTβR) (64, 65), or Id2 (66) have revealed important roles for each of these

4 Feb 2002

12:2

AR

AR152-24.tex

AR152-24.SGM

LaTeX2e(2001/05/10)

P1: IKH

Annu. Rev. Immunol. 2002.20:795-823. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

RANK-L AND RANK

803

molecules in the development and organization of secondary lymphoid tissues. For example, TNFα activation of the TNF-R1 is required for the formation of splenic B lymphocyte follicles, follicular dendritic networks, and germinal center formation (63, 67). Mice with disrupted LTα, LTβ, or LTβR genes lack lymph nodes, Peyer’s patches, and follicular dendritic cells; such mice show altered splenic architecture (59, 61, 62, 65). Thus, it was assumed that lymph node organogenesis and the development of Peyer’s patches are always genetically linked. Surprisingly, rankl−/− and rank−/− mice lack all lymph nodes but display intact splenic architecture and develop Peyer’s patches normally, suggesting that RANK-L and RANK have a specific and essential role in lymph node organogenesis. Importantly, RANK-L disruption provided the first evidence that development of lymph nodes and Peyer’s patches can be genetically uncoupled. The concerted activity of several cell lineages including fibroblasts, macrophages, reticular cells, and endothelial cells is required for the morphogenesis of primordial lymph nodes (8). These primordial lymph nodes are subsequently seeded by T and B cells and CD4+CD3−LTβ + cells that differentiate into NK cells, antigen presenting cells, and follicular cells to form mature compact nodes (68). In situ hybridization of normal lymph nodes has shown that RANK-Land RANK-expressing cells are present in lymph nodes, located mainly in the cortical areas adjacent to subcapsular sinuses (38). The identity of these cells has yet to be determined. Since RANK and RANK-L are also expressed in the spleen and Peyer’s patches, restricted RANK-L-RANK expression cannot account for the selective lack of lymph nodes. Moreover, because defective homing of rankl−/− lymphocytes was excluded as the cause of defective lymph node formation and normal bone marrow cells cannot rescue the lymph node defect in rankl−/− mice in chimeric transfer experiments, we speculated that RANK-L may act as a growth and/or survival factor on a lymph node organizing cell during embryonic development (58). Recently it has been shown that the defective lymph node development in rankl−/− mice correlates with a significant reduction in lymphotoxin LTαβ +α4β7+CD45+CD4+CD3− cells and their failure to form clusters in rudimentary mesenteric (69). Transgenic RANK-L-mediated restoration of lymph node development required LTαβ expression on CD45+CD4+CD3− cells as lymph node formation could not be induced in LTα −/− mice. The authors proposed that both RANK-L and LTαβ regulate the colonization and cluster formation by CD45+CD4+CD3− cells during lymph node organogenesis (69). Similar to rankl−/− and rank−/− mice, TNF-R1−/− mice exhibit retained but small Peyer’s patches (63, 67), which suggests that both TNF-R1 and RANK-L may have a potential, albeit not essential, role in the formation of Peyer’s patches. We have recently generated rankl-tnfr1 double knockout mice, and these mice completely lack Peyer’s patches in the small intestine without further affecting the defects in splenic architecture observed in tnfr1 single mutant mice (Y. Kong and J. M. Penninger, unpublished). Thus, RANK-L is essential for lymph node formation and cooperates with the TNFR1 in the formation of Peyer’s patches. The exact cellular and molecular mechanisms of RANK-L-RANK-regulated lymph node

8 Feb 2002

10:50

804

AR

THEILL

AR152-24.tex

¥

BOYLE

¥

AR152-24.SGM

LaTeX2e(2001/05/10)

P1: IKH

PENNINGER

morphogenesis and the linkage between lymph node and Peyer’s patch formation need to be tested.

Annu. Rev. Immunol. 2002.20:795-823. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

RANK-L-RANK and Dendritic Cells—In Search of a Function? Similar to the CD40L/CD40 system, interactions between RANK-L expressed on activated T cells and RANK expressed on dendritic cells can mediate DC survival via Bcl-XL induction and upregulation of the costimulatory molecule CD40 on DCs (Figure 4) (12, 56, 57). Recently it has also been shown that RANK-L activates the anti-apoptotic serine/threonine kinase Akt/PKB through a signaling complex involving TRAF6 and c-Src on mature DCs and osteoclasts (35). In addition to Akt/PKB activation, NF-κB and ERK are activated by RANK-L. Because NKκB, ERK, and Akt/PKB promote cell survival by inhibiting apoptosis-inducing pathways, activation of these anti-apoptotic molecules seems to be at least partially responsible for the RANK-L-mediated DC survival. In addition to these in vitro studies, it has been shown that treatment of antigen-pulsed mature DCs with soluble RANK-L in vitro enhances the number and persistence of antigenpresenting DCs in the draining lymph nodes in vivo (56). Furthermore, RANK-L treatment increased antigen-specific primary T cell responses. Interestingly, significant memory responses were observed only in mice injected with RANK-Ltreated DCs (56). The increase in primary and memory T cell responses following vaccination with RANK-L-treated DCs could be due to enhanced/altered cytokine production such as expression of IL-12 and/or an increased number of antigenpulsed DCs. Both CD40L and RANK-L have functional similarity, are expressed on activated T cells, and enhance the activation and survival of DCs (12, 57). However, in contrast to CD40L/CD40, RANK-L/RANK signaling does not alter the expression of cell surface molecules such as MHC class II, CD80, CD86, and CD54. Whereas CD40L is primarily expressed on activated CD4+ T cells, RANK-L is expressed on activated CD4+ and CD8+ T cells (70, 17). Moreover, the maximal level of RANK-L following the intial T cell activation event occurs at 48 h, and high levels of RANK-L expression are sustained until 96 h, while CD40L is rapidly expressed and downregulated (71). Thus, CD40L-CD40 interactions may primarily control the initial priming stage, whereas RANK-L-RANK may act at later times than does CD40L during the immune response. For example, CD40L is essential for the T cell-dependent B cell responses such as germinal center formation, affinity maturation, and class switching (72, 73). By contrast, in rankl−/− mice, germinal center formation, Ig class-switching, and the production of neutralizing antiviral Abs are not overtly affected, and all the B cell defects could be explained by the absence of lymph nodes (Y.-Y. Kong and J. M. Penninger, unpublished). Inhibition of RANK-L in vivo using a soluble RANK-Fc molecule does not block the priming of LCMV-specific T cells, but it does impair proliferation of CD4+ T cells to the viral antigen at later time points after infection (74). This impaired CD4+ T cell response in RANK-Fc treated mice was especially apparent

4 Feb 2002

12:2

AR

AR152-24.tex

AR152-24.SGM

LaTeX2e(2001/05/10)

P1: IKH

Annu. Rev. Immunol. 2002.20:795-823. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

RANK-L AND RANK

805

in the absence of CD40-expression. Thus, at later stages of the immune response, RANK-L can regulate CD40L-independent activation of CD4+ T helper cells (74). These observations suggest that although CD40L and RANK-L have functional similarity and may cooperate, RANK-L and CD40-L may also have fundamentally different functions in the control of immune responses: CD40L regulates T/B responses, and RANK-L appears to have a role in memory T cell responses. However, in our laboratory using OPG transgenic mice and bone marrow chimeric mice that have normal lymph nodes but lack RANK-L expression on lymphocytes, we noted modulation of RANK-L does not impair the generation of cytototoxic T cells (P. Ohashi, J. N. Penninger, unpublished). Only activated T cells, but not resting T cells, express RANK-L, which promotes dendritic cell survival (57). DCs reside in tissues as immature cells and are specialized to capture and process antigens that lead to maturation of DCs in response to inflammatory stimuli. Mature DCs that have captured antigens migrate to T cell zones of secondary lymphoid organs by afferent lymphatics in order to present antigen to antigen-specific T cells. The T cell areas of secondary lymphoid organs represent the microenvironment that allows interactions between DCs, T cells, and B cells to initiate adaptive immune responses (reviewed in 75). Antigenbearing DCs are in direct contact with na¨ıve antigen-specific T cells within the T cell areas of lymph nodes, and after interaction with T cells these DCs are eliminated rapidly (76). Activated T cells induce apoptosis of DCs by producing the TNF-family molecules TRAIL, FasL, and TNFα. Accumulation and prolonged survival of DCs were reported in patients with human autoimmune lymphoproliferative syndrome type II. These patients had a caspase-10 mutation that rendered DCs resistant to TRAIL-induced cell death (77). Thus, it appears that mature DCs have short life-spans and that mature DCs presenting antigens to T cells must be effectively eliminated to avoid excessive immune responses. The life span of DCs may be an important checkpoint to control for the induction of tolerance, priming, and chronic inflammation (78). Since both TRAIL and RANK-L are produced by activated T cells, the balance between RANK-L and TRAIL may influence DC survival (79–81). Both RANK-L and TRAIL can bind to OPG (53) and OPG is made by DCs (44), which suggests that these factors control the fate of DCs. Based on these studies, various groups are currently trying to control the DC fate via RANK-L-RANK and OPG to modulate in vivo DC survival and to enhance the efficiency of DC-based vaccinations for anti-tumor therapy or the treatment of autoimmune diseases. In the final analyses of all the published genetic and functional studies on RANK-L, RANK, and OPG, it appears that although these molecules can influence some aspects of lymphocyte and DC functions, none of these molecules plays an essential function in T cells, B cells, or DCs that cannot be compensated for by other molecules such as CD40L/CD40. Thus, the essential and true function of RANKL/RANK in the immune system and communication between DCs and T cells need to be elucidated. For example, since expression of these molecules can be controlled by sex hormones (54, 82),

4 Feb 2002

12:2

806

AR

AR152-24.tex

THEILL

¥

BOYLE

AR152-24.SGM

¥

LaTeX2e(2001/05/10)

P1: IKH

PENNINGER

we speculate that this system may control gender specific differences in immunity and could be involved in the higher incidence of autoimmune diseases like arthritis in women.

Annu. Rev. Immunol. 2002.20:795-823. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

Lymphocyte Differentiation Two principal genetic checkpoints regulate thymocyte differentiation. The first checkpoint, at the CD44−CD25+ stage of development, depends on the expression of the pre-TCR on CD4−CD8− thymocyte precursors, which regulates expansion of these precursor cells. The second checkpoint regulates progression from CD4+CD8+ immature to mature CD4+ or CD8+ thymocytes and correlates with positive thymocyte selection. Various mutations that arrest thymocyte development at the stage of pre-TCR expression have been reported (83). All of these mutations either affect the pre-TCR complex directly or affect signaling molecules thought to be downstream of the pre-TCR. RANK-L expression has been detected on CD4−CD8− early thymocyte precursors (12). RANK-L-deficient mice showed the block in the progression of CD4−CD8−CD44−CD25+ precursors to CD4−CD8−CD44−CD25− thymocytes. This developmental defect does not reside in the thymic environment but is intrinsic to bone marrow–derived cells. These data suggested that the TNF-family cytokine RANK-L is important for the progression of CD25+CD44− precursors to CD25−CD44− thymocytes at the stage of pre-TCR expression. However, thymuses of newborn rankl−/− mice and differentiation of rankl−/− thymocytes in fetal thymic organ cultures appear normal, and the defects of thymocyte development in rankl−/− mice are only apparent at around 2 weeks of age, a phenotype reminiscent of thymic defects in mice lacking the pro-apoptotic bcl-2 family molecule bim (84). Age-dependent interactions between RANK-Lexpressing thymocyte precursors and as-yet-unknown thymic stromal cells expressing RANK could contribute to early thymocyte development and thymocyte expansion, whereas later stages appear to be RANK-L independent. However, although RANK mRNA can be found in the thymus using in situ hybridization (58), rank−/− mice do not display any obvious defects in thymocyte maturation (23, 24). This difference in thymocyte differentiation is the only discernable distinction between rankl and rank mutant mice, and it suggests that RANK-L might act on another, yet unidentified, receptor during early thymocyte development. Additional work in the future should provide evidence of whether RANK-L and RANK indeed play a role in thymocyte development in a cell-autonomous fashion. For example, since RANK-L and OPG expression can be controlled by sex hormones, it would be interesting to test whether the thymocyte differentiation defects observed in rankl−/−, but not rank−/− mice, are dependent on sex hormone levels, a scenario that could explain age-related differences. In addition to T cells, rankl and rank knockout mice have reduced numbers of mature B220+IgD+ and B220+IgM+ B cells in the spleen and lymph nodes and slightly disorganized B cell areas in primary splenic follicles (20, 23, 58). Since

4 Feb 2002

12:2

AR

AR152-24.tex

AR152-24.SGM

LaTeX2e(2001/05/10)

P1: IKH

Annu. Rev. Immunol. 2002.20:795-823. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

RANK-L AND RANK

807

rankl and rank-null mice have no bone marrow cavities, the reduced cellularity of B cells could be due to an altered microenvironment or to changes in the composition of stromal cells outside the bone marrow cavity that affect B cell differentiation. For example, rankl−/− mice form an ectopically organized extramedullary hematopoietic tissue localized at the outer surfaces of vertebral bodies (58). This tissue exhibits morphological and phenotypic features characteristic of hematopoiesis and proliferating precursor cells. Whether these hematopoietic islands in rankl−/− mice represent a defect in the homing of precursors during the switch from hepatic to bone marrow haematopoiesis, or an event secondary to osteopetrosis that interferes with the seeding of bone marrow cavities, remains to be determined. In fetal liver cell chimeras, RANK-L regulates early B cell differentiation from the B220+CD43+CD25− pro-B cell to the B220+CD43−CD25+ pre-B cell stage of development, which indicates that the TNF-family cytokine RANK-L is indeed a regulator of early B lymphocyte development (58). Recent evidence in a new opg mutant mouse strain confirms the notion that the interplay of RANK-L-RANK and the molecular decoy receptor OPG may regulate the development and possibly the function of B lymphocytes (85). Ex vivo, opg−/− pro-B cells have enhanced proliferation to IL-7, and type 1 transitional B cells accumulate in the spleens of opg−/− mice. Thus, loss of OPG may control B cell maturation. Moreover, it should be noted that OPG is a CD40-regulated gene in B cells and dendritic cells (85) and that prostaglandin E2 treatment can increase the amount of RANK-L messenger RNA in B220+ B cells in an estrogen-dependent manner (86). Whether RANK-L acts as a survival factor required for early B and T cell development or whether RANK-L directly affects antigen receptor-driven lymphocyte maturation also remains to be seen.

T Cells and Bone—The Emergence of Osteoimmunology Bone remodeling and bone loss are controlled by a balance between RANKL/RANK and the RANK-L decoy receptor OPG. Since RANK-L is made by T cells following antigen-receptor stimulation, we asked the question whether T cell–derived RANK-L can indeed regulate the development and activation of osteoclasts, that is, whether activated T cells can modulate bone turnover via RANK-L. In an in vitro cell culture system, activated T cells can directly trigger osteoclastogenesis via RANK-L (17). Importantly, systemic activation of T cells in vivo leads to a RANK-L-dependent increase in osteoclastogenesis followed by bone loss. All in vitro and in vivo effects of T cells on osteoclasts could be blocked by the administration of the decoy receptor OPG (17). Moreover, in a recent elegant study it has been shown that transgenic overexpression of RANK-L in T cells restores osteoclastogenesis in a rankl−/− background and partially restores normal bone marrow cavities (87). These data showed that systemic activation of T cells leads to bone loss, indicating that, through their production of RANK-L, T cells are crucial mediators of bone loss in vivo. The results also provided a novel paradigm for T cells as regulators of bone physiology.

4 Feb 2002

12:2

Annu. Rev. Immunol. 2002.20:795-823. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

808

AR

AR152-24.tex

THEILL

¥

BOYLE

AR152-24.SGM

¥

LaTeX2e(2001/05/10)

P1: IKH

PENNINGER

Because mutant mice that lack T cells still have normal bone cavities and tooth eruption, T cells are probably not required for normal bone homeostasis. However, chronic systemic T cell activation such as in autoimmune diseases, viral infections, or local inflammation within the bone due to metastasis, infections, and fractures, or joint inflammation in arthritis all probably attract T cells that then actively participate in bone remodeling via production of RANK-L. Moreover, glucocorticoids, which are used to treat autoimmune diseases and allergic disorders, strongly induce RANK-L expression and decrease OPG (15, 88). Interestingly, our own unpublished results indicate that glucocorticoids and the TCR induce RANK-L expression in T cells. Thus, in certain diseases such as asthma, primary activation of T cells together with immunosuppressive treatment may in fact exacerbate osteopenia via synergistic activation of RANK-L expression on T cells. These findings provide a molecular explanation for bone loss associated with diseases having immune system involvement, such as adult and childhood leukemias, cancer metastasis, autoimmunity, and various viral infections. Inhibition of RANK-L function via OPG or a related molecule may therefore prevent bone destruction in multiple diseases.

RANK-L is a Critical Mediator of Crippling in Arthritis One disease in which osteoclasts are a critical factor for disease progression is arthritis; bone loss results in life-long crippling. Arthritis in humans is characterized by synovial inflammation, erosion of bone and cartilage, severe joint pain and ultimately life-long crippling (89). In Lewis rats, experimental induction of arthritis by subcutaneous injection of bacterial products in adjuvant leads to severe inflammation in the bone marrow and soft tissues surrounding joints, accompanied by extensive local bone and cartilage destruction, loss of bone mineral density, and crippling (90). This condition in rats, called adjuvant-induced arthritis (AdA), mimics many of the clinical and pathological features of human RA. Lesions in rat AdA are dependent on T cell activation (91), and T cells in the inflamed joints and draining lymph nodes of affected rats produce many pro-inflammatory cytokines (89). In this model system of severe arthritis that mimics many of the clinical and pathological features of human RA, RANK-L protein is expressed on the surface of synovial effector T cells isolated at the clinical onset of arthritis (17). Inhibition of RANK-L via OPG had no effect on the severity of inflammation. However, OPG treatment completely abolished the loss of mineral bone density in the inflamed joints of these animals, in a dose-dependent manner. Histologically, OPG-treated arthritic rats exhibited minimal loss of cortical and trabecular bone, whereas untreated arthritic animals developed severe bone lesions characterized by partial to complete destruction of cortical and trabecular bone, and erosion of the articular cartilages. Bone destruction in untreated arthritic rats correlated with a dramatic increase in osteoclast numbers, whereas OPG treatment prevented the accumulation of osteoclasts (17). These results showed that RANK-L is a key mediator of

4 Feb 2002

12:2

AR

AR152-24.tex

AR152-24.SGM

LaTeX2e(2001/05/10)

P1: IKH

Annu. Rev. Immunol. 2002.20:795-823. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

RANK-L AND RANK

809

joint destruction and bone loss in adjuvant arthritis. Importantly, whereas untreated rats experienced severe crippling, rats treated with OPG at the onset of disease— similar to a patient consulting a doctor at the onset of joint swelling—did not show any signs of clinical crippling. Alteration of cartilage structures leading to cartilage collapse constitutes a critical step in arthritic joint destruction. Controversy exists whether cartilage destruction occurs independently of bone loss, or whether damage to the subchondral bone indirectly causes cartilage deterioration (92). In untreated arthritic rats, partial or complete erosion of the cartilage in both the central and peripheral regions of joint surfaces is observed. In striking contrast, the integrity of cartilage was preserved in OPG-treated arthritic rats. Neither cartilage erosion nor matrix degeneration in the centers of joint surfaces occurred in OPG-treated animals (17). OPG could protect the cartilage by maintaining the underlying subchondral bone and insulating the overlying cartilage from the inflammatory cell infiltrates in the bone marrow. Since both RANK-L and RANK are expressed on chondrocytes (13, 20), and rankl (58) as well as rank mutant mice (24) exhibit significant changes in the columnar alignment of chondrocytes at the growth plate, it is possible that RANK-L/RANK play a direct role in cartilage growth and cartilage homeostasis. These data provided the first evidence that inhibition of RANK-L activity by OPG can prevent cartilage destruction, a critical, irreversible step in the pathogenesis of arthritis. It has become evident that the development of arthritis can occur in the absence of T cells (93). Using in situ hybridization of inflamed rat joints and isolation of different cell populations from these joints, we could show that RANK-L is indeed expressed in lymphocytes, macrophages, and especially in synoviocytes (17). In line with these findings, genetic ablation of RANK-L also does not prevent inflammation in an antibody-mediated model of arthritis using the K/BxN serum transfer model (94). Multinucleated TRAP-positive osteoclast-like cells were abundant in resorption lacunae in areas of bone erosion in arthritic control mice, and they were completely absent in arthritic TRANCE/RANK-L knockout mice, demonstrating the absolute requirement for TRANCE/RANK-L in osteoclastogenesis in this serum transfer model of inflammatory arthritis (95). Cartilage damage was still observed in both arthritic TRANCE/RANK-L knockout mice and arthritic control mice, but a trend toward milder cartilage damage in the TRANCE/RANK-L knockout mice was noted. Thus, TRANCE/RANK-L apparently is not required for cartilage destruction, but clearly plays an as-yet-unidentified modulatory role (95). Moreover, inhibition of RANK-L via OPG prevents bone loss without affecting inflammation in a TNFα-induced arthritis model (96), indicating that TNFα triggered bone loss is critically dependent on RANK-L expression (S. Smolen, Vienna, personal communication). Whether OPG prevents bone in other animal models of arthritis needs to be tested. To investigate whether RANK-L is implicated in human RA, we collected inflammatory cells from the synovial fluid of patients with adult or juvenile rheumatoid arthritis (RA) and patients with osteoarthritis, and we evaluated OPG and RANK-L expression. All RA and osteoarthritis patients ever tested (n > 40)

8 Feb 2002

10:53

810

AR

THEILL

AR152-24.tex

¥

BOYLE

¥

AR152-24.SGM

LaTeX2e(2001/05/10)

P1: IKH

PENNINGER

Annu. Rev. Immunol. 2002.20:795-823. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

exhibited RANK-L expression in inflammatory cells, whereas OPG expression was not detectable (17). Thus, the correlation between RANK-L expression in inflamed joints and arthritis appears to be absolute. To distinguish which cells were producing RANK-L, inflammatory synovial fluids were separated into T and nonT cell populations. Consistent with our results in rats, both synovial T and non-T cell populations from RA patients expressed RANK-L, but not OPG. Fibroblasts isolated from the same adult RA joints failed to express RANK-L. These data confirm the findings in rodent adjuvant arthritis and suggest that RANK-L is the principal mediator of bone destruction in human arthritis.

A Molecular Scenario of T Cell–Regulated Bone Loss In inflammatory or autoimmune disease states, activated T cells produce RANK-L and pro-inflammatory cytokines such as TNFα, IL-1, or IL-11, all of which can induce RANK-L expression in osteoblasts and bone marrow stromal cells (97). Thus, it appears that T cells promote bone resorption directly via RANK-L expression and indirectly via expression of pro-inflammatory cytokines that mediate RANKL expression in non-T cells (Figure 5). These results are in line with the findings that T cells and non-T cell populations express RANK-L in arthritic joints. Inhibition of RANK-L has no effect on inflammation but completely prevents bone loss and protects cartilage (17). Bone resorption induced by local injection of IL-1β or TNFα over the calvaria of mice can be blocked by concurrent systemic injection of OPG, which indicates that RANK-L is the mediator of the bone-damaging effects of TNFα and IL1-β in vivo (98). Although inhibition of TNFα and IL-1 using soluble receptor antagonists to some extent prevents inflammation and bone loss in arthritis (7, 99, 100), inhibition of RANK-L function via OPG might therefore prevent bone destruction and cartilage damage in arthritis irrespective of the initial trigger. Reduced bone mineral densities can also be seen in many human diseases such as adult and childhood leukemia (101), chronic infections such as hepatitis C or HIV (102), autoimmune disorders such as diabetes mellitus (103) and lupus erythematosus (104), allergic diseases such as asthma (105), lytic bone metastases in multiple cancers such as breast cancer (106), and of course arthritis (89). These osteopenic disorders can cause irreversible crippling, thereby severely disrupting the lives of significant numbers of patients. For example, many patients with lupus require hip replacement surgery, and essentially all children that survive leukemia experience severe bone loss and growth retardation. In North America and Europe, 1 in 100 people develop RA and 1 in 10 people develop osteoarthritis. In addition, T cell–derived RANK-L contributes to alveolar bone resorption and tooth loss in an animal model that mimics human periodontal disease. The alveolar bone resorption around the teeth can be inhibited with OPG (107). Moreover, it should be noted that OPG can prevent cancer cell–induced bone destruction (108–110), cancer metastases–associated bone pain, and painrelated neurochemical reorganization of the spinal chord (111). In most of these

8 Feb 2002

10:53

AR

AR152-24.tex

AR152-24.SGM

LaTeX2e(2001/05/10)

P1: IKH

Annu. Rev. Immunol. 2002.20:795-823. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

RANK-L AND RANK

811

osteopenic disorders, disease pathogenesis correlates with the activation of T cells (Figure 5). However, short-term activation of T cells does not result in any detectable bone loss, not even in some chronic T cell and TNFα-mediated diseases such as ankylosis spondylitis (112). Moreover, T cells are working constantly to fight off the universe of foreign particles in which we live, so, at any point in time, some T cells are activated (113). What is it that prevents these T cells from causing extensive bone loss every time we have an infection? A crucial counter-regulatory mechanism, by which activated T cells can inhibit the RANK-L-induced maturation and activation of osteoclasts has been recently discovered (Figure 3) (114). It turns out that interferon-γ blocks RANK-L-induced osteoclast differentiation in vitro. Moreover, interferon-γ receptor knockout mice develop more bone destruction in inflammatory arthritis than do normal mice. Mechanistically, interferon-γ can activate the ubiquitin-proteasome pathway within the osteoclasts, resulting in the degradation of TRAF6. Thus, it appears that interferon-γ can prevent uncontrolled bone loss during inflammatory T cell responses. Moreover, T cell–derived IL-12 alone, and IL-12 in synergy with IL-18, inhibits osteoclast formation in vitro (115), and IL-4 can abrogate osteoclastogenesis through STAT6-dependent inhibition of NF-κB signaling (116, 117). Thus, multiple T cell–derived cytokines might be able to interfere with RANK signaling and therefore with osteoclastogenesis and osteoclast functions. In the future it will be interesting to determine the mechanisms that control the balance between T cell–mediated bone loss and inhibition of osteoclastogenesis. Nonetheless it has become clear now that inhibition of RANKL-mediated activation of RANK via OPG or a related molecule ameliorates many osteopenic conditions. RANK-L inhibition appears to be the most rational and advisable strategy to prevent bone destruction in multiple diseases, to possibly eradicate major human diseases such as osteoporosis, to curtail crippling, and to limit tooth loss, diseases that affect millions of people.

Bone Loss, Mammary Gland Formation, and Mammalian Evolution The expression of RANK-L and OPG is regulated by multiple hormones and cytokines shown to affect the development and activation of osteoclasts, including 25-dihydroxyvitamin D3, IL-1, IL-11, PGE2, calcitonin, and TNFα (11). Intriguingly, expression of RANK-L and OPG is also strongly influenced by the female sex hormones progesterone and estrogen, and by hormones involved in reproduction and lactation such as prolactin and parathyroid hormone-related peptide (PTHrP) (118). Reduction of ovarian function following menopause in women, and ovariectomy in animal models, result in osteoporosis and fractures, conditions that can be completely reversed at least in animals by treatment with OPG (38). However, the evolutionary and functional rationale for RANK-L/OPG regulation by reproductive hormones and the prevalence of hormonally regulated and gender-biased osteoporosis in older females were not known.

4 Feb 2002

12:2

Annu. Rev. Immunol. 2002.20:795-823. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

812

AR

AR152-24.tex

THEILL

¥

BOYLE

AR152-24.SGM

¥

LaTeX2e(2001/05/10)

P1: IKH

PENNINGER

In mammals, sex and pregnancy hormones control mammary gland morphogenesis and formation of a lactating mammary gland. Mammary gland morphogenesis proceeds in distinct steps, beginning with a fetal mammary anlage that undergoes ductal elongation and branching (119). During pregnancy, increased ductal side branching and development of lobulo-alveolar structures result from the expansion and proliferation of ductal and alveolar epithelium (120). Formation of a lactating mammary gland provides essential nourishment to mammalian newborns in the form of milk. Calcium is an important component of milk, and the main source of calcium for a newborn mammal is its mother’s breast milk (121). Calcium transport from mothers to the fetus and neonates is a vital process to preserve species. Deficits in maternal calcium transfer or calcium handling in the offspring have severe consequences for newborns, ranging from rickets to heart and brain defects (122). Mothers meet the increased requirements for calcium during pregnancy and lactation by doubling their intestinal calcium absorption and demineralizing their skeletons via activation of bone-resorbing osteoclasts (123). Surprisingly, mice lacking RANK-L or its receptor RANK fail to form lobuloalveolar mammary gland structures during pregnancy and show a complete block in the formation of a lactating mammary gland, leading to the death of newborn pups. RANK-L expression in mammary epithelial cells is induced by pregnancy hormones, whereas the RANK is constitutively expressed on these cells. Transplantation and local RANK-L-rescue experiments in rankl−/− and rank−/− pregnant females showed that RANK-L acts directly on RANK-expressing mammary epithelial cells. The effects of RANK-L are autonomous to epithelial cells. The mammary gland defect in female rankl−/− mice is characterized by enhanced apoptosis and by failures in proliferation and Akt/PKB activation in lobulo-alveolar buds that can be reversed by recombinant RANK-L treatment. Thus, RANK-L and RANK, the master regulators of skeletal calcium release, are essential for the formation of the lactating mammary gland, the organ required for transmission of maternal calcium to neonates in mammalian species. Importantly, these data provided a novel function for TNF and TNFR family proteins and a new paradigm in the formation of a lactating mammary gland. In phylogenetic evolution, the formation of lactating mammary glands is a relatively recent event, occurring when the first mammals appeared about 200 million years ago. Thus, mammals took a gene product that is the master regulator of bone metabolism and calcium turnover in the whole organism and subverted it to stimulation of mammary gland development during pregnancy. The balance between RANK-L, RANK, and the decoy receptor OPG is critical for the regulation of bone loss in osteoporosis, arthritis, and lytic bone metastases (118). Osteoporosis affects hundreds of millions of people, particularly postmenopausal women. Intriguingly, genetic and functional models have shown that osteoclast-regulated bone remodeling is under the control of powerful sex and pregnancy hormones (11). When estrogen production falls, such as occurs in ovariectomy models in animals or in postmenopausal women, induction of OPG is decreased, allowing uncontrolled demineralization of bone by RANK-L-stimulated osteoclasts (38). Osteoporosis is

4 Feb 2002

12:2

AR

AR152-24.tex

AR152-24.SGM

LaTeX2e(2001/05/10)

P1: IKH

Annu. Rev. Immunol. 2002.20:795-823. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

RANK-L AND RANK

813

strongly associated with increased morbidity and premature death in older women. The evolutionary question then arises, why has the RANK-L/RANK/OPG system, a key regulator of a structural organ such as the skeleton, come under the regulatory umbrella of reproductive hormones? Our results provided an unexpected molecular and evolutionary explanation for gender bias and the high incidence of osteoporosis in females. The strong bias toward bone loss in postmenopausal women may be due to the fact that the RANK-L/RANK/OPG system is essential for reproduction and the survival of mammalian offspring. The transcription of the decoy receptor OPG is also regulated by pregnancy hormones, particularly estrogen (54). Thus, local or systemic OPG may contribute to mammary gland formation during pregnancy. However, opg−/− females (40) and females from three different OPG-overexpressing transgenic lines (38) showed normal breeding and mammary gland formation (our own unpublished data). Moreover, implantation of soluble OPG pellets into wild-type pregnant females at day 13.5 of pregnancy did not affect any aspect of mammary gland development (J. Tata, Y. Y. Kong, J. M. Penninger, unpublished). Thus, high systemic or local levels of OPG apparently cannot inhibit RANK-L and RANK interaction, implying that RANK-L acts locally on the same or neighboring epithelial cells. The action of OPG and the connection of RANK-L/RANK with pregnancy hormones also provide further insights into the regulation of bone loss during pregnancy. The binding of RANK-L to RANK on mature osteoclasts triggers their activation that then leads to release of calcium from the skeleton. OPG competes with RANK-L for binding to RANK and thus is a potent inhibitor of osteoclast differentiation and activity. Since estrogen enhances OPG expression on osteoblasts, we suggest that increased estrogen levels may protect the maternal skeleton via OPG during pregnancy. After giving birth, maternal estrogen levels rapidly decline, leading to decreased OPG and permitting the mobilization of calcium from the bones for lactation. Although this hypothesis explains clinical and experimental observations in animal models and humans, e.g., calcium-release from the bone due to osteoclasts activation is predominant in the third trimester of gestation and during lactating (121, 123), it awaits proper testing. During osteoclast development, CSF-1 and RANK-L cooperate to stimulate the differentiation of hematopoietic progenitors into mature multinucleated osteoclasts (13). CSF-1 provides the survival signal, whereas RANK-L is the critical factor for osteoclast lineage determination. In contrast to the synergy between CSF-1 and RANK-L in osteoclastogenesis, CSF-1 appears to act independently of RANK-L/ RANK during mammary gland formation. Unlike rankl−/− and rank−/− mice, csf-1−/− mice show increased lobulo-alveolar development of mammary epithelium during pregnancy and ovarian defects (124). Thus, it appears that RANK-L/ RANK has a specific and unique role in mammary gland development that is regulated by pregnancy hormones. Mice deficient for the stat5a, cyclin D1, or prolactin receptor genes have defects in mammary gland development similar to those observed in rankl−/− and rank−/− females (119, 125, 126), which suggests that RANK-L/RANK, Cyclin D1, Stat5a and/or prolactin might cooperate to

8 Feb 2002

10:55

Annu. Rev. Immunol. 2002.20:795-823. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

814

AR

THEILL

AR152-24.tex

¥

BOYLE

¥

AR152-24.SGM

LaTeX2e(2001/05/10)

P1: IKH

PENNINGER

stimulate lobulo-alveolar development. The exact functional and genetic relationships between RANK-L/RANK and pregnancy hormones, signaling molecules, and cell cycle regulatory molecules await elucidation. These data have another intriguing implication. Both RANK-L and RANK are required for lymph node organogenesis, and RANK-L expressed on T cells provides activation and survival signals to dendritic cells. DCs, which express OPG, are specialized antigen-presenting cells that initiate and integrate immune responses. Like calcium uptake, immune responses must be modified during pregnancy so that a mother does not reject her allogeneic fetus, although the same mother can still reject an allogeneic skin transplant. Other examples of immune system alterations include the observed amelioration of clinical symptoms of multiple sclerosis during pregnancy, and the onset of severe T cell–regulated food allergies in some pregnant women (127, 128). Thus, pregnancy is associated with alterations to the immune system that do not impair the general response to infections but lead to selective immunological adjustments (127). Moreover, like osteoporosis, the development of autoimmunity shows a gender bias, and various sex/pregnancy hormones such as estrogen and prolactin influence the function and development of various lymphocyte populations (129, 130). Since RANK-L, RANK, and OPG provide a genetic interface between the immune system, bone remodeling, and formation of a lactating mammary gland, this system is an intriguing starting point to address these questions at the genetic level.

CONCLUSIONS RANK-L, its receptor RANK, and the decoy receptor OPG are the key regulators for osteoclast development and the activation of mature osteoclasts (Figure 5). Surprisingly, the same molecules that regulate osteoclastogenesis were identified as a key factor in early differentiation of thymocytes and B cell precursors and the development of lymph nodes and Peyer’s patches. In the immune system, RANKL is produced by activated T cells and acts as a potent survival factor of DCs. The understanding and manipulation of DC fate by RANK-L/RANK provides a new avenue for anti-tumor vaccination and the treatment of autoimmune diseases. RANK-L produced by activated T cells can directly induce osteoclastogenesis. Systemic or local activation of T cells triggers bone loss via expression of RANK-L. These findings provide the molecular explanation for bone loss associated with diseases having immune system involvement, such as adult and childhood leukemias, autoimmunity, and various viral infections. Inhibition of RANK-L function via OPG or a related molecule might therefore ameliorate many osteopenic conditions and prevent bone destruction and cartilage damage that ultimately cause crippling in arthritis. Moreover, RANK-L and RANK, the master regulators of skeletal calcium release, are essential for the morphogenesis of a lactating mammary gland, and they provide an evolutionary rationale for hormonal regulation of osteoporosis.

4 Feb 2002

12:2

AR

AR152-24.tex

AR152-24.SGM

LaTeX2e(2001/05/10)

P1: IKH

RANK-L AND RANK

815

ACKNOWLEDGMENTS The authors are supported by Amgen Inc. J. M. P. holds a Canadian Research Chair in Cell Biology and is supported by the Premiers Research Excellence Award, The National Cancer Institute of Canada (NCIC), and the Canadian Institute for Health Research (CIHR). We thank all the members of our laboratories for vital contributions and discussion.

Annu. Rev. Immunol. 2002.20:795-823. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

Visit the Annual Reviews home page at www.annualreviews.org

LITERATURE CITED 1. Carswell EA, Old LJ, Kassel RL, Green S, Fiore N, Williamson B. 1975. An endotoxin-induced serum factor that causes necrosis of tumors. Proc. Natl. Acad. Sci. USA 72:3666–70 2. Granger GA, Shacks SJ, Williams TW, Kolb WP. 1969. Lymphocyte in vitro cytotoxicity: specific release of lymphotoxinlike materials from tuberculin-sensitive lymphoid cells. Nature 221:1155–57 3. Wallach D. 1996. Suicide by order: some open questions about the cell-killing activities of the TNF ligand and receptor families. Cytokine Growth Factor Rev. 7:211–21 4. Locksley RM, Killeen N, Lenardo MJ. 2001. The TNF and TNF receptor superfamilies: integrating mammalian biology. Cell 104:487–501 5. Strasser A, Newton K. 1999. FADD/ MORT1, a signal transducer that can promote cell death or cell growth. Int. J. Biochem. Cell Biol. 31:533–37 6. Yamada Y, Kirillova I, Peschon JJ, Fausto N. 1997. Initiation of liver growth by tumor necrosis factor: deficient liver regeneration in mice lacking type I tumor necrosis factor receptor. Proc. Natl. Acad. Sci. USA 94:1441–46 7. Williams RO, Feldmann M, Maini RN. 2000. Cartilage destruction and bone erosion in arthritis: the role of tumour necrosis factor alpha. Ann. Rheum. Dis. 59:75– 80 8. Fu YX, Chaplin DD. 1999. Development

9.

10. 11.

12.

13.

14.

and maturation of secondary lymphoid tissues. Annu. Rev. Immunol. 17:399–433 Grewal IS, Flavell RA. 1998. CD40 and CD154 in cell-mediated immunity. Annu. Rev. Immunol. 16:111–35 Roodman GD. 1999. Cell biology of the osteoclast. Exp. Hematol. 27:1229–41 Ross FP. 2000. RANKing the importance of measles virus in Paget’s disease. J. Clin. Invest. 105:555–58 Anderson DM, Maraskovsky E, Billingsley WL, Dougall WC, Tometsko ME, Roux ER, Teepe MC, DuBose RF, Cosman D, Galibert L. 1997. A homologue of the TNF receptor and its ligand enhance T-cell growth and dendritic-cell function. Nature 390:175–79 Lacey DL, Timms E, Tan HL, Kelley MJ, Dunstan CR, Burgess T, Elliott R, Colombero A, Elliott G, Scully S, Hsu H, Sullivan J, Hawkins N, Davy E, Capparelli C, Eli A, Qian YX, Kaufman S, Sarosi I, Shalhoub V, Senaldi G, Guo J, Delaney J, Boyle WJ. 1998. Osteoprotegerin ligand is a cytokine that regulates osteoclast differentiation and activation. Cell 93:165–76 Wong BR, Rho J, Arron J, Robinson E, Orlinick J, Chao M, Kalachikov S, Cayani E, Bartlett FS 3rd, Frankel WN, Lee SY, Choi Y. 1997. TRANCE is a novel ligand of the tumor necrosis factor receptor family that activates c-Jun Nterminal kinase in T cells. J. Biol. Chem. 272:25190–94

4 Feb 2002

12:2

Annu. Rev. Immunol. 2002.20:795-823. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

816

AR

AR152-24.tex

THEILL

¥

BOYLE

AR152-24.SGM

¥

LaTeX2e(2001/05/10)

P1: IKH

PENNINGER

15. Yasuda H, Shima N, Nakagawa N, Yamaguchi K, Kinosaki M, Mochizuki S, Tomoyasu A, Yano K, Goto M, Murakami A, Tsuda E, Morinaga T, Higashio K, Udagawa N, Takahashi N, Suda T. 1998. Osteoclast differentiation factor is a ligand for osteoprotegerin/osteoclastogenesis-inhibitory factor and is identical to TRANCE/RANKL. Proc. Natl. Acad. Sci. USA 95:3597–602 16. Lum L, Wong BR, Josien R, Becherer JD, Erdjument-Bromage H, Schlondorff J, Tempst P, Choi Y, Blobel CP. 1999. Evidence for a role of a tumor necrosis factoralpha (TNF-alpha)-converting enzymelike protease in shedding of TRANCE, a TNF family member involved in osteoclastogenesis and dendritic cell survival. J. Biol. Chem. 274:13,613–18 17. Kong YY, Feige U, Sarosi I, Bolon B, Tafuri A, Morony S, Capparelli C, Li J, Elliott R, McCabe S, Wong T, Campagnuolo G, Moran E, Bogoch ER, Van G, Nguyen LT, Ohashi PS, Lacey DL, Fish E, Boyle WJ, Penninger JM. 1999. Activated T cells regulate bone loss and joint destruction in adjuvant arthritis through osteoprotegerin ligand. Nature 402:304– 9 18. Nakashima T, Kobayashi Y, Yamasaki S, Kawakami A, Eguchi K, Sasaki H, Sakai H. 2000. Protein expression and functional difference of membrane-bound and soluble receptor activator of NFkappaB ligand: modulation of the expression by osteotropic factors and cytokines. Biochem. Biophys. Res. Commun. 275:768–75 19. Lagasse E, Weissman IL. 1997. Enforced expression of Bcl-2 in monocytes rescues macrophages and partially reverses osteopetrosis in op/op mice. Cell 89:1021– 31 20. Hsu H, Lacey DL, Dunstan CR, Solovyev I, Colombero A, Timms E, Tan HL, Elliott G, Kelley MJ, Sarosi I, Wang L, Xia XZ, Elliott R, Chiu L, Black T, Scully S, Capparelli C, Morony S, Shima-

21.

22.

23.

24.

25.

26.

moto G, Bass MB, Boyle WJ. 1999. Tumor necrosis factor receptor family member RANK mediates osteoclast differentiation and activation induced by osteoprotegerin ligand. Proc. Natl. Acad. Sci. USA 96:3540–45 Burgess TL, Qian Y, Kaufman S, Ring BD, Van G, Capparelli C, Kelley M, Hsu H, Boyle WJ, Dunstan CR, Hu S, Lacey DL. 1999. The ligand for osteoprotegerin (OPGL) directly activates mature osteoclasts. J. Cell Biol. 145:527–38 Nakagawa N, Kinosaki M, Yamaguchi K, Shima N, Yasuda H, Yano K, Morinaga T, Higashio K. 1998. RANK is the essential signaling receptor for osteoclast differentiation factor in osteoclastogenesis. Biochem. Biophys. Res. Commun. 253:395–400 Dougall WC, Glaccum M, Charrier K, Rohrbach K, Brasel K, De Smedt T, Daro E, Smith J, Tometsko ME, Maliszewski CR, Armstrong A, Shen V, Bain S, Cosman D, Anderson D, Morrissey PJ, Peschon JJ, Schuh J. 1999. RANK is essential for osteoclast and lymph node development. Genes Dev. 13:2412–24 Li J, Sarosi I, Yan XQ, Morony S, Capparelli C, Tan HL, McCabe S, Elliott R, Scully S, Van G, Kaufman S, Juan SC, Sun Y, Tarpley J, Martin L, Christensen K, McCabe J, Kostenuik P, Hsu H, Fletcher F, Dunstan CR, Lacey DL, Boyle WJ. 2000. RANK is the intrinsic hematopoietic cell surface receptor that controls osteoclastogenesis and regulation of bone mass and calcium metabolism. Proc. Natl. Acad. Sci. USA 97:1566–71 Hughes AE, Ralston SH, Marken J, Bell C, MacPherson H, Wallace RGH, van H-W, Whyte MP, Nakatsuka K, Hovy L, Anderson DM. 2000. Mutations in TNFRSF11A, affecting the signal peptide of RANK, cause familial expansile osteolysis. Nat. Gen. 24:45–48 Darnay BG, Haridas V, Ni J, Moore PA, Aggarwal BB. 1998. Characterization of the intracellular domain of receptor

4 Feb 2002

12:2

AR

AR152-24.tex

AR152-24.SGM

LaTeX2e(2001/05/10)

P1: IKH

RANK-L AND RANK

Annu. Rev. Immunol. 2002.20:795-823. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

27.

28.

29.

30.

31.

32.

33.

activator of NF-kappaB (RANK). Interaction with tumor necrosis factor receptorassociated factors and activation of NFkappab and c-Jun N-terminal kinase. J. Biol. Chem. 273:20,551–55 Galibert L, Tometsko ME, Anderson DM, Cosman D, Dougall WC. 1998. The involvement of multiple tumor necrosis factor receptor (TNFR)-associated factors in the signaling mechanisms of receptor activator of NF-kappaB, a member of the TNFR superfamily. J. Biol. Chem. 273:34,120–27 Lee ZH, Kwack K, Kim KK, Lee SH, Kim HH. 2000. Activation of c-Jun Nterminal kinase and activator protein 1 by receptor activator of nuclear factor kappaB. Mol. Pharmacol. 58:1536–45 Wong BR, Josien R, Lee SY, Vologodskaia M, Steinman RM, Choi Y. 1998. The TRAF family of signal transducers mediates NF-kappaB activation by the TRANCE receptor. J. Biol. Chem. 273: 28355–59 Kobayashi N, Kadono Y, Naito A, Matsumoto K, Yamamoto T, Tanaka S, Inoue J. 2001. Segregation of TRAF6-mediated signaling pathways clarifies its role in osteoclastogenesis. Embo J. 20:1271–80 Lomaga MA, Yeh WC, Sarosi I, Duncan GS, Furlonger C, Ho A, Morony S, Capparelli C, Van G, Kaufman S, van der Heiden A, Itie A, Wakeham A, Khoo W, Sasaki T, Cao Z, Penninger JM, Paige CJ, Lacey DL, Dunstan CR, Boyle WJ, Goeddel DV, Mak TW. 1999. TRAF6 deficiency results in osteopetrosis and defective interleukin-1, CD40, and LPS signaling. Genes Dev. 13:1015–24 Naito A, Azuma S, Tanaka S, Miyazaki T, Takaki S, Takatsu K, Nakao K, Nakamura K, Katsuki M, Yamamoto T, Inoue J. 1999. Severe osteopetrosis, defective interleukin-1 signalling and lymph node organogenesis in TRAF6-deficient mice. Genes Cells 4:353–62 Franzoso G, Carlson L, Xing LP, Poljak L, Shores EW, Brown KD, Leonardi A,

34.

35.

36.

37.

38.

39.

40.

817

Tran T, Boyce BF, Siebenlist U. 1997. Requirement for NF-kappa B in osteoclast and B-cell development. Genes Dev. 11:3482–96 Iotsova V, Caamano J, Loy J, Yang Y, Lewin A, Bravo R. 1997. Osteopetrosis in mice lacking NF-kappa B1 and NF-kappa B2. Nat. Med. 3:1285–89 Wong BR, Besser D, Kim N, Arron JR, Vologodskaia M, Hanafusa H, Choi Y. 1999. TRANCE, a TNF family member, activates Akt/PKB through a signaling complex involving TRAF6 and c-Src. Mol. Cell 4:1041–49 Arron JR, Vologodskaia M, Wong BR, Naramura M, Kim N, Gu H, Choi Y. 2001. A positive regulatory role for Cbl family proteins in TRANCE and CD40Lmediated Akt activation. J. Biol. Chem. 13:13 Matsumoto M, Sudo T, Saito T, Osada H, Tsujimoto M. 2000. Involvement of p38 mitogen-activated protein kinase signaling pathway in osteoclastogenesis mediated by receptor activator of NF-kappa B ligand (RANKL). J. Biol. Chem. 275: 31155–61 Simonet WS, Lacey DL, Dunstan CR, Kelley M, Chang MS, Luthy R, Nguyen HQ, Wooden S, Bennett L, Boone T, Shimamoto G, DeRose M, Elliott R, Colombero A, Tan HL, Trail G, Sullivan J, Davy E, Bucay N, Renshaw-Gegg L, Hughes TM, Hill D, Pattison W, Campbell P, Boyle WJ, et al. 1997. Osteoprotegerin: a novel secreted protein involved in the regulation of bone density. Cell 89: 309–19 Yasuda H, Shima N, Nakagawa N, Mochizuki SI, Yano K, Fujise N, Sato Y, Goto M, Yamaguchi K, Kuriyama M, Kanno T, Murakami A, Tsuda E, Morinaga T, Higashio K. 1998. Identity of osteoclastogenesis inhibitory factor (OCIF) and osteoprotegerin (OPG): a mechanism by which OPG/OCIF inhibits osteoclastogenesis in vitro. Endocrinology 139:1329–37 Bucay N, Sarosi I, Dunstan CR, Morony

4 Feb 2002

12:2

818

Annu. Rev. Immunol. 2002.20:795-823. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

41.

42.

43.

44.

45.

46.

47.

AR

AR152-24.tex

THEILL

¥

BOYLE

AR152-24.SGM

¥

LaTeX2e(2001/05/10)

P1: IKH

PENNINGER

S, Tarpley J, Capparelli C, Scully S, Tan HL, Xu W, Lacey DL, Boyle WJ, Simonet WS. 1998. Osteoprotegerin-deficient mice develop early onset osteoporosis and arterial calcification. Genes Dev. 12: 1260–68 Mizuno A, Murakami A, Nakagawa N, Yasuda H, Tsuda E, Morinaga T, Higashio K. 1998. Structure of the mouse osteoclastogenesis inhibitory factor (OCIF) gene and its expression in embryogenesis. Gene 215:339–43 Brandstrom H, Jonsson KB, Ohlsson C, Vidal O, Ljunghall S, Ljunggren O. 1998. Regulation of osteoprotegerin mRNA levels by prostaglandin E2 in human bone marrow stroma cells. Biochem. Biophys. Res. Commun. 247:338–41 Brandstrom H, Jonsson KB, Vidal O, Ljunghall S, Ohlsson C, Ljunggren O. 1998. Tumor necrosis factor-alpha and -beta upregulate the levels of osteoprotegerin mRNA in human osteosarcoma MG-63 cells. Biochem. Biophys. Res. Commun. 248:454–57 Yun TJ, Chaudhary PM, Shu GL, Frazer JK, Ewings MK, Schwartz SM, Pascual V, Hood LE, Clark EA. 1998. OPG/ FDCR-1, a TNF receptor family member, is expressed in lymphoid cells and is upregulated by ligating CD40. J. Immunol. 161:6113–21 Min H, Morony S, Sarosi I, Dunstan CR, Capparelli C, Scully S, Van G, Kaufman S, Kostenuik PJ, Lacey DL, Boyle WJ, Simonet WS. 2000. Osteoprotegerin reverses osteoporosis by inhibiting endosteal osteoclasts and prevents vascular calcification by blocking a process resembling osteoclastogenesis. J. Exp. Med. 192:463–74 Parhami F, Demer LL. 1997. Arterial calcification in face of osteoporosis in ageing: Can we blame oxidized lipids? Curr. Opin. Lipidol. 8:312–14 Jorgensen L, Engstad T, Jacobsen BK. 2001. Bone mineral density in acute stroke patients: low bone mineral density may

48.

49.

50.

51.

52.

53.

54.

55.

predict first stroke in women. Stroke 32:47–51 Walker-Bone K, Dennison E, Cooper C. 2001. Epidemiology of osteoporosis. Rheum. Dis. Clin. North Am. 27:1–18 Mach F, Schonbeck U, Sukhova GK, Atkinson E, Libby P. 1998. Reduction of atherosclerosis in mice by inhibition of CD40 signalling. Nature 394:200–3 Azuma Y, Kaji K, Katogi R, Takeshita S, Kudo A. 2000. Tumor necrosis factoralpha induces differentiation of and bone resorption by osteoclasts. J. Biol. Chem. 275:4858–64 Kobayashi K, Takahashi N, Jimi E, Udagawa N, Takami M, Kotake S, Nakagawa N, Kinosaki M, Yamaguchi K, Shima N, Yasuda H, Morinaga T, Higashio K, Martin TJ, Suda T. 2000. Tumor necrosis factor alpha stimulates osteoclast differentiation by a mechanism independent of the ODF/RANKL-RANK interaction. J. Exp. Med. 191:275–85 Zhang YH, Heulsmann A, Tondravi MM, Mukherjee A, Abu-Amer Y. 2001. Tumor necrosis factor-alpha (TNF) stimulates RANKL-induced osteoclastogenesis via coupling of TNF type 1 receptor and RANK signaling pathways. J. Biol. Chem. 276:563–68 Emery JG, McDonnell P, Burke MB, Deen KC, Lyn S, Silverman C, Dul E, Appelbaum ER, Eichman C, DiPrinzio R, Dodds RA, James IE, Rosenberg M, Lee JC, Young PR. 1998. Osteoprotegerin is a receptor for the cytotoxic ligand TRAIL. J. Biol. Chem. 273:14363–67 Hofbauer LC, Khosla S, Dunstan CR, Lacey DL, Spelsberg TC, Riggs BL. 1999. Estrogen stimulates gene expression and protein production of osteoprotegerin in human osteoblastic cells. Endocrinology 0:4367 Saika M, Inoue D, Kido S, Matsumoto T. 2001. 17beta-estradiol stimulates expression of osteoprotegerin by a mouse stromal cell line, ST-2, via estrogen receptoralpha. Endocrinology 142:2205–2

4 Feb 2002

12:2

AR

AR152-24.tex

AR152-24.SGM

LaTeX2e(2001/05/10)

P1: IKH

Annu. Rev. Immunol. 2002.20:795-823. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

RANK-L AND RANK 56. Josien BR, Li HL, Ingulli E, Sarma S, B RW, Vologodskaia M, Steinman RM, Choi Y. 2000. TRANCE, a tumor necrosis factor family member, enhances the longevity and adjuvant properties of dendritic cells In vivo. J. Exp. Med. 191:495– 502 57. Wong BR, Josien R, Lee SY, Sauter B, Li HL, Steinman RM, Choi Y. 1997. TRANCE (tumor necrosis factor [TNF]related activation-induced cytokine), a new TNF family member predominantly expressed in T cells, is a dendritic cellspecific survival factor. J. Exp. Med. 186: 2075 58. Kong YY, Yoshida H, Sarosi I, Tan HL, Timms E, Capparelli C, Morony S, Oliveira-dos-Santos AJ, Van G, Itie A, Khoo W, Wakeham A, Dunstan CR, Lacey DL, Mak TW, Boyle WJ, Penninger JM. 1999. OPGL is a key regulator of osteoclastogenesis, lymphocyte development and lymph-node organogenesis. Nature 397:315–23 59. De Togni P, Goellner J, Ruddle NH, Streeter PR, Fick A, Mariathasan S, Smith SC, Carlson R, Shornick LP, Strauss-Schoenberger J, et al. 1994. Abnormal development of peripheral lymphoid organs in mice deficient in lymphotoxin. Science 264:703 60. Rennert PD, Browning JL, Hochman PS. 1997. Selective disruption of lymphotoxin ligands reveals a novel set of mucosal lymph nodes and unique effects on lymph node cellular organization. Int. Immunol. 9:1627–39 61. Alimzhanov MB, Kuprash DV, KoscoVilbois MH, Luz A, Turetskaya RL, Tarakhovsky A, Rajewsky K, Nedospasov SA, Pfeffer K. 1997. Abnormal development of secondary lymphoid tissues in lymphotoxin beta-deficient mice. Proc. Natl. Acad. Sci. USA 94:9302 62. Koni PA, Sacca R, Lawton P, Browning JL, Ruddle NH, Flavell RA. 1997. Distinct roles in lymphoid organogenesis for lymphotoxins alpha and beta revealed in

63.

64.

65.

66.

67.

68.

69.

70.

819

lymphotoxin beta-deficient mice. Immunity 6:491–500 Matsumoto M, Mariathasan S, Nahm MH, Baranyay F, Peschon JJ, Chaplin DD. 1996. Role of lymphotoxin and the type I TNF receptor in the formation of germinal centers. Science 271:1289 Futterer A, Mink K, Luz A, Kosco-Vilbois MH, Pfeffer K. 1998. The lymphotoxin beta receptor controls organogenesis and affinity maturation in peripheral lymphoid tissues. Immunity 9:59–70 Rennert PD, James D, Mackay F, Browning JL, Hochman PS. 1998. Lymph node genesis is induced by signaling through the lymphotoxin beta receptor. Immunity 9:71–79 Yokota Y, Mansouri A, Mori S, Sugawara S, Adachi S, Nishikawa S, Gruss P. 1999. Development of peripheral lymphoid organs and natural killer cells depends on the helix-loop-helix inhibitor Id2. Nature 397:702–6 Pasparakis M, Alexopoulou L, Grell M, Pfizenmaier K, Bluethmann H, Kollias G. 1997. Peyer’s patch organogenesis is intact yet formation of B lymphocyte follicles is defective in peripheral lymphoid organs of mice deficient for tumor necrosis factor and its 55-kDa receptor. Proc. Natl. Acad. Sci. USA 94:6319–23 Mebius RE, Rennert P, Weissman IL. 1997. Developing lymph nodes collect CD4+CD3-LTbeta+ cells that can differentiate to APC, NK cells, and follicular cells but not T or B cells. Immunity 7:493– 504 Kim D, Mebius RE, MacMicking JD, Jung S, Cupedo T, Castellanos Y, Rho J, Wong BR, Josien R, Kim N, Rennert PD, Choi Y. 2000. Regulation of peripheral lymph node genesis by the tumor necrosis factor family member TRANCE. J. Exp. Med. 192:1467–78 Josien R, Wong BR, Li HL, Steinman RM, Choi Y. 1999. TRANCE, a TNF family member, is differentially expressed on T cell subsets and induces cytokine

4 Feb 2002

12:2

820

71.

Annu. Rev. Immunol. 2002.20:795-823. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

72.

73.

74.

75.

76.

77.

78.

79.

AR

AR152-24.tex

THEILL

¥

BOYLE

AR152-24.SGM

¥

LaTeX2e(2001/05/10)

P1: IKH

PENNINGER

production in dendritic cells. J. Immunol. 162:2562–68 Roy M, Waldschmidt T, Aruffo A, Ledbetter JA, Noelle RJ. 1993. The regulation of the expression of gp39, the CD40 ligand, on normal and cloned CD4+ T cells. J. Immunol. 151:2497–2510 Kawabe T, Naka T, Yoshida K, Tanaka T, Fujiwara H, Suematsu S, Yoshida N, Kishimoto T, Kikutani H. 1994. The immune responses in CD40-deficient mice: impaired immunoglobulin class switching and germinal center formation. Immunity 1:167–78 Xu J, Foy TM, Laman JD, Elliott EA, Dunn JJ, Waldschmidt TJ, Elsemore J, Noelle RJ, Flavell RA. 1994. Mice deficient for the CD40 ligand. Immunity 1:423–31 Bachmann MF, Wong BR, Josien R, Steinman RM, Oxenius A, Choi Y. 1999. TRANCE, a tumor necrosis factor family member critical for CD40 ligandindependent T helper cell activation. J. Exp. Med. 189:1025–31 Banchereau J, Steinman RM. 1998. Dendritic cells and the control of immunity. Nature 392:245–52 Ingulli E, Mondino A, Khoruts A, Jenkins MK. 1997. In vivo detection of dendritic cell antigen presentation to CD4(+) T cells. J. Exp. Med. 185:2133–41 Wang J, Zheng L, Lobito A, Chan FK, Dale J, Sneller M, Yao X, Puck JM, Straus SE, Lenardo MJ. 1999. Inherited human Caspase 10 mutations underlie defective lymphocyte and dendritic cell apoptosis in autoimmune lymphoproliferative syndrome type II. Cell 98:47–58 Sauter B, Albert ML, Francisco L, Larsson M, Somersan S, Bhardwaj N. 2000. Consequences of cell death. Exposure to necrotic tumor cells, but not primary tissue cells or apoptotic cells, induces the maturation of immunostimulatory dendritic cells. J. Exp. Med. 191:423–34 Kayagaki N, Yamaguchi N, Nakayama M, Kawasaki A, Akiba H, Okumura

80.

81.

82.

83.

84.

85.

86.

K, Yagita H. 1999. Involvement of TNFrelated apoptosis-inducing ligand in human CD4+ T cell-mediated cytotoxicity. J. Immunol. 162:2639–47 Martinez-Lorenzo MJ, Anel A, Gamen S, Monlen I, Lasierra P, Larrad L, Pineiro A, Alava MA, Naval J. 1999. Activated human T cells release bioactive Fas ligand and APO2 ligand in microvesicles. J. Immunol. 163:1274–81 Thomas WD, Hersey P. 1998. TNF-related apoptosis-inducing ligand (TRAIL) induces apoptosis in Fas ligand-resistant melanoma cells and mediates CD4 T cell killing of target cells. J. Immunol. 161: 2195–200 Fata JE, Kong YY, Li J, Sasaki T, IrieSasaki J, Moorehead RA, Elliott R, Scully S, Voura EB, Lacey DL, Boyle WJ, Khokha R, Penninger JM. 2000. The osteoclast differentiation factor osteoprotegerin-ligand is essential for mammary gland development. Cell 103:41–50 von Boehmer H, Fehling HJ. 1997. Structure and function of the pre-T cell receptor. Annu. Rev. Immunol. 15:433–52 Bouillet P, Metcalf D, Huang DCS, Tarlinton DM, Kay TWH, Kontgen F, Adams JM, Strasser A. 1999. Proapoptotic Bcl-2 relative bim required for certain apoptotic responses, leukocyte homeostasis, and to preclude autoimmunity. Science 286:1735–38 Yun TJ, Tallquist MD, Aicher A, Rafferty KL, Marshall AJ, Moon JJ, Ewings ME, Mohaupt M, Herring SW, Clark EA. 2001. Osteoprotegerin, a crucial regulator of bone metabolism, also regulates B cell development and function. J. Immunol. 166:1482–91 Kanematsu M, Sato T, Takai H, Watanabe K, Ikeda K, Yamada Y. 2000. Prostaglandin E2 induces expression of receptor activator of nuclear factor-kappa B ligand/osteoprotegrin ligand on pre-B cells: implications for accelerated osteoclastogenesis in estrogen deficiency. J. Bone Miner. Res. 15:1321–29

4 Feb 2002

12:2

AR

AR152-24.tex

AR152-24.SGM

LaTeX2e(2001/05/10)

P1: IKH

Annu. Rev. Immunol. 2002.20:795-823. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

RANK-L AND RANK 87. Kim N, Odgren PR, Kim DK, Marks SC Jr, Choi Y. 2000. Diverse roles of the tumor necrosis factor family member TRANCE in skeletal physiology revealed by TRANCE deficiency and partial rescue by a lymphocyte-expressed TRANCE transgene. Proc. Natl. Acad. Sci. USA 97:10,905–10 88. Vidal NO, Brandstrom H, Jonsson KB, Ohlsson C. 1998. Osteoprotegerin mRNA is expressed in primary human osteoblastlike cells: down-regulation by glucocorticoids. J. Endocrinol. 159:191–95 89. Feldmann M, Brennan FM, Maini RN. 1996. Role of cytokines in rheumatoid arthritis. Annu. Rev. Immunol. 14:397– 440 90. Bendele A, McComb J, Gould T, McAbee T, Sennello G, Chlipala E, Guy M. 1999. Animal models of arthritis: relevance to human disease. Toxicol. Pathol. 27:134–42 91. Panayi GS, Lanchbury JS, Kingsley GH. 1992. The importance of the T cell in initiating and maintaining the chronic synovitis of rheumatoid arthritis. Arth. Rheum. 35:729–35 92. Conway JG, Wakefield JA, Brown RH, Marron BE, Sekut L, Stimpson SA, McElroy A, Menius JA, Jeffreys JJ, Clark RL, McGeehan GM, Connolly KM. 1995. Inhibition of cartilage and bone destruction in adjuvant arthritis in the rat by a matrix metalloproteinase inhibitor. J. Exp. Med. 182:449–57 93. Benoist C, Mathis D. 2000. A revival of the B cell paradigm for rheumatoid arthritis pathogenesis? Arthritis Res. 2:90–94 94. Kouskoff V, Korganow AS, Duchatelle V, Degott C, Benoist C, Mathis D. 1996. Organ-specific disease provoked by systemic autoimmunity. Cell 87:811–22 95. Pettit AR, Ji H, von Stechow D, M¨uller R, Choi Y, Goldring SR, Benoist C, Gravallese EM. 2001. TRANCE/RANKL knockout mice are protected from bone erosion in the K/BxN serum transfer model of arthritis. Am. J. Pathol. In press

821

96. Keffer J, Probert L, Cazlaris H, Georgopoulos S, Kaslaris E, Kioussis D, Kollias G. 1991. Transgenic mice expressing human tumour necrosis factor: a predictive genetic model of arthritis. EMBO J. 10:4025–31 97. Hofbauer LC, Lacey DL, Dunstan CR, Spelsberg TC, Riggs BL, Khosla S. 1999. Interleukin-1beta and tumor necrosis factor-alpha, but not interleukin-6, stimulate osteoprotegerin ligand gene expression in human osteoblastic cells. Bone 25:255–59 98. Morony S, Capparelli C, Lee R, Shimamoto G, Boone T, Lacey DL, Dunstan CR. 1999. A chimeric form of osteoprotegerin inhibits hypercalcemia and bone resorption induced by IL-1beta, TNF-alpha, PTH, PTHrP, and 1, 25(OH)2D3. J. Bone Miner. Res. 14:1478–85 99. Fye KH. 1999. New treatments for rheumatoid arthritis. Available and upcoming slow-acting antirheumatic drugs. Postgrad. Med. 106:82–85 100. Graninger WB, Smolen JS. 2001. Oneyear inhibition of tumor necrosis factoralpha: a major success or a larger puzzle? Curr. Opin. Rheumatol. 13:9–13 101. Oliveri MB, Mautalen CA, Rodriguez Fuchs CA, Romanelli MC. 1991. Vertebral compression fractures at the onset of acute lymphoblastic leukemia in a child. Henry Ford Hosp. Med. J. 39:45–48 102. Stellon AJ, Davies A, Compston J, Williams R. 1985. Bone loss in autoimmune chronic active hepatitis on maintenance corticosteroid therapy. Gastroenterology 89:1078–83 103. Piepkorn B, Kann P, Forst T, Andreas J, Pfutzner A, Beyer J. 1997. Bone mineral density and bone metabolism in diabetes mellitus. Hormone Metab. Res. 29:584– 91 104. Seitz M, Hunstein W. 1985. Enhanced prostanoid release from monocytes of patients with rheumatoid arthritis and active systemic lupus erythematosus. Ann. Rheum. Dis. 44:438–45

4 Feb 2002

12:2

Annu. Rev. Immunol. 2002.20:795-823. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

822

AR

AR152-24.tex

THEILL

¥

BOYLE

AR152-24.SGM

¥

LaTeX2e(2001/05/10)

P1: IKH

PENNINGER

105. Ebeling PR, Erbas B, Hopper JL, Wark JD, Rubinfeld AR. 1998. Bone mineral density and bone turnover in asthmatics treated with long-term inhaled or oral glucocorticoids. J. Bone Miner. Res. 13: 1283–89 106. Coleman RE. 1998. How can we improve the treatment of bone metastases further? Curr. Opin. Oncol. 10 (Suppl.)1:7– 13 107. Teng YT, Nguyen H, Gao X, Kong YY, Gorczynski RM, Singh B, Ellen RP, Penninger JM. 2000. Functional human T-cell immunity and osteoprotegerin ligand control alveolar bone destruction in periodontal infection. J. Clin. Invest. 106:59– 67 108. Atkins GJ, Bouralexis S, Haynes DR, Graves SE, Geary SM, Evdokiou A, Zannettino AC, Hay S, Findlay DM. 2001. Osteoprotegerin inhibits osteoclast formation and bone resorbing activity in giant cell tumors of bone. Bone 28:370– 77 109. Morony S, Capparelli C, Sarosi I, Lacey DL, Dunstan CR, Kostenuik PJ. 2001. Osteoprotegerin inhibits osteolysis and decreases skeletal tumor burden in syngeneic and nude mouse models of experimental bone metastasis. Cancer Res. 61:4432–36 110. Zhang J, Dai J, Qi Y, Lin DL, Smith P, Strayhorn C, Mizokami A, Fu Z, Westman J, Keller ET. 2001. Osteoprotegerin inhibits prostate cancer-induced osteoclastogenesis and prevents prostate tumor growth in the bone. J. Clin. Invest. 107:1235–44 111. Honore P, Luger NM, Sabino MAC, Schwei MJ, Rogers SD, Mach DB, O’Keefe PF, Ramnaraine ML, Clohisy DR, Mantyh PW. 2000. Osteoprotegerin blocks bone cancer-induced skeletal destruction, skeletal pain and pain-related neurochemical reorganization of the spinal cord. Nat. Med. 6:521–28 112. Brandt J, Haibel H, Cornely D, Golder W, Gonzalez J, Reddig J, Thriene W,

113. 114.

115.

116.

117.

118.

119.

120.

121.

122.

Sieper J, Braun J. 2000. Successful treatment of active ankylosing spondylitis with the anti-tumor necrosis factor alpha monoclonal antibody infliximab. Arthritis Rheum. 43:1346–52 Arron JR, Choi Y. 2000. Bone versus immune system. Nature 408:535–36 Takayanagi H, Ogasawara K, Hida S, Chiba T, Murata S, Sato K, Takaoka A, Yokochi T, Oda H, Tanaka K, Nakamura K, Taniguchi T. 2000. T-cell-mediated regulation of osteoclastogenesis by signalling cross-talk between RANKL and IFN-gamma. Nature 408:600–5 Horwood NJ, Elliott J, Martin TJ, Gillespie MT. 2001. IL-12 alone and in synergy with IL-18 inhibits osteoclast formation in vitro. J. Immunol. 166:4915–21 Abu-Amer Y. 2001. IL-4 abrogates osteoclastogenesis through STAT6-dependent inhibition of NF-kappaB. J. Clin. Invest. 107:1375–85 Bendixen AC, Shevde NK, Dienger KM, Willson TM, Funk CD, Pike JW. 2001. IL-4 inhibits osteoclast formation through a direct action on osteoclast precursors via peroxisome proliferator-activated receptor gamma 1. Proc. Natl. Acad. Sci. USA 98:2443–48 Karsenty G. 1999. The genetic transformation of bone biology. Genes Dev. 13:3037–51 Robinson GW, Karpf AB, Kratochwil K. 1999. Regulation of mammary gland development by tissue interaction. J. Mamm. Gland Biol. Neoplasia 4:9–19 Robinson GW, Hennighausen L, Johnson PF. 2000. Side-branching in the mammary gland: the progesterone-Wnt connection. Genes Dev. 14:889–94 Kovacs CS, Kronenberg HM. 1997. Maternal-fetal calcium and bone metabolism during pregnancy, puerperium, and lactation. Endocrinol. Rev. 18:832– 72 Pereira GR, Zucker AH. 1986. Nutritional deficiencies in the neonate. Clin. Perinatol. 13:175–89

4 Feb 2002

12:2

AR

AR152-24.tex

AR152-24.SGM

LaTeX2e(2001/05/10)

P1: IKH

Annu. Rev. Immunol. 2002.20:795-823. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

RANK-L AND RANK 123. Cross NA, Hillman LS, Allen SH, Krause GF, Vieira NE. 1995. Calcium homeostasis and bone metabolism during pregnancy, lactation, and postweaning: a longitudinal study. Am. J. Clin. Nutr. 61:514– 23 124. Pollard JW, Hennighausen L. 1994. Colony stimulating factor 1 is required for mammary gland development during pregnancy. Proc. Natl. Acad. Sci. USA 91:9312–16 125. Brisken C, Kaur S, Chavarria TE, Binart N, Sutherland RL, Weinberg RA, Kelly PA, Ormandy CJ. 1999. Prolactin controls mammary gland development via direct and indirect mechanisms. Dev. Biol. 210:96–106 126. Fantl V, Edwards PA, Steel JH, Vonderhaar BK, Dickson C. 1999. Impaired mammary gland development in Cyl-

127.

128.

129.

130.

823

1(−/−) mice during pregnancy and lactation is epithelial cell autonomous. Dev. Biol. 212:1–11 Buyon JP. 1998. The effects of pregnancy on autoimmune diseases. J. Leuk. Biol. 63:281–87 Whitacre CC, Reingold SC, O’Looney PA. 1999. A gender gap in autoimmunity. Science 283:1277–78 Ahmed SA, Hissong BD, Verthelyi D, Donner K, Becker K, Karpuzoglu-Sahin E. 1999. Gender and risk of autoimmune diseases: possible role of estrogenic compounds. Environ. Health Perspect. 107: 681–86 Rider V, Abdou NI. 2001. Gender differences in autoimmunity: molecular basis for estrogen effects in systemic lupus erythematosus. Int. Immunopharmacol. 1:1009–24

Annu. Rev. Immunol. 2002.20:795-823. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

6 Feb 2002

8:22

AR

AR152-24-COLOR.tex

AR152-24-COLOR.SGM

LaTeX2e(2002/01/18)

Figure 1 Osteoclast lineage development. Mutations that affect osteoclastogenesis and activation of mature osteoclasts are indicated. For details see text.

P1: GDL

Annu. Rev. Immunol. 2002.20:795-823. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

6 Feb 2002

8:22

AR

AR152-24-COLOR.tex

AR152-24-COLOR.SGM

LaTeX2e(2002/01/18)

P1: GDL

Figure 2 Regulation of osteoclast formation in bone tissues. Calciotropic factors such as vitamin-D3, prostaglandin E2, IL-1, IL-11, TNFα and glucocorticoid induce RANKL expression on osteoblasts. RANKL binding to the RANK expressed on haematopoietic progenitors activates a signal transduction cascade that leads to osteoclast differentiation in the presence of the survival factor CSF-1. Moreover, RANKL stimulates bone resorbing activity in mature osteoclasts via RANK. OPG produced by osteoblasts acts as a decoy receptor for RANKL and inhibits osteoclastogenesis and osteoclast activation by binding to RANKL. TGFβ released from bone during active bone resorption has been suggested as a feedback mechanism by upregulating OPG level. Estrogen can enhance OPG production on osteoblasts which is a possible explanation of postmenopausal osteoporosis following estrogen withdrawal.

Annu. Rev. Immunol. 2002.20:795-823. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

6 Feb 2002

8:22

AR

AR152-24-COLOR.tex

AR152-24-COLOR.SGM

LaTeX2e(2002/01/18)

P1: GDL

Figure 3 RANK signaling pathways. When RANK is activated it sends signals into the cells through tumor necrosis factor receptor-associated factors (TRAFs) 1, 2, 3, 5, and 6. In addition, c-Src and Cbl proteins associate with the cytoplasmatic tail of RANK. These RANK-associated molecules relay RANK-dependent stimulation to downstream pathways such as NF-kB, JNK/SAPK, p38, and Akt/PKB that regulate bone resorption, activation, survival, and differentiation of osteoclasts and dendritic cells. IFNγ can inhibit RANKL-mediated osteoclastogenesis presumably via induction of TRAF6 ubiquitination and proteolytic TRAF6 degradation. The scheme is based on ref. (113).

Annu. Rev. Immunol. 2002.20:795-823. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

6 Feb 2002

8:22

AR

AR152-24-COLOR.tex

AR152-24-COLOR.SGM

LaTeX2e(2002/01/18)

P1: GDL

Figure 4 RANKL promotes DC survival via binding to RANK. This function is similar to that of CD40L-CD40 interactions. TRAIL produced by activated T cells and DCs induces apoptosis of DCs. The balance between RANKL and TRAIL levels at late stages of immune responses might determine the fate of DCs. Manipulation of DC fate via TRAIL, CD40L, RANKL, and OPG, could be used to optimize DC-based anti-tumor vaccination protocols and to treat autoimmune diseases. It should be noted that DC development and in vitro functions for T-cell activation appear normal in rankl and rank mutant mice. The functional capacity and longevity of rankl and rank mutant DCs in vivo needs to be addressed.

Annu. Rev. Immunol. 2002.20:795-823. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

6 Feb 2002

8:22

AR

AR152-24-COLOR.tex

AR152-24-COLOR.SGM

LaTeX2e(2002/01/18)

P1: GDL

Figure 5 Activated T cells affect bone physiology. Activated T cells produce cytokines such as TNFα, IL-1, IL-11, and IL-17 that lead to RANKL expression on osteoblasts. Moreover, activated T cells directly express and produce RANKL that induces osteoclast formation and activation. The soluble decoy receptor for RANKL, OPG, blocks both pathways. Inhibition of RANKL via OPG might be useful to treat osteoporosis, crippling in arthritis, osteopenic disorders such as Paget’s disease, and bone loss and pain associated with bone metastases.

P1: FRK

February 1, 2002

12:47

Annual Reviews

AR152-FM

Annual Review of Immunology Volume 20, 2002

Annu. Rev. Immunol. 2002.20:795-823. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

CONTENTS FRONTISPIECE, Charles A. Janeway, Jr. A TRIP THROUGH MY LIFE WITH AN IMMUNOLOGICAL THEME, Charles A. Janeway, Jr.

xiv 1

THE B7 FAMILY OF LIGANDS AND ITS RECEPTORS: NEW PATHWAYS FOR COSTIMULATION AND INHIBITION OF IMMUNE RESPONSES, Beatriz M. Carreno and Mary Collins

29

MAP KINASES IN THE IMMUNE RESPONSE, Chen Dong, Roger J. Davis, and Richard A. Flavell

PROSPECTS FOR VACCINE PROTECTION AGAINST HIV-1 INFECTION AND AIDS, Norman L. Letvin, Dan H. Barouch, and David C. Montefiori T CELL RESPONSE IN EXPERIMENTAL AUTOIMMUNE ENCEPHALOMYELITIS (EAE): ROLE OF SELF AND CROSS-REACTIVE ANTIGENS IN SHAPING, TUNING, AND REGULATING THE AUTOPATHOGENIC T CELL REPERTOIRE, Vijay K. Kuchroo, Ana C. Anderson, Hanspeter Waldner, Markus Munder, Estelle Bettelli, and Lindsay B. Nicholson

55 73

101

NEUROENDOCRINE REGULATION OF IMMUNITY, Jeanette I. Webster, Leonardo Tonelli, and Esther M. Sternberg

125

MOLECULAR MECHANISM OF CLASS SWITCH RECOMBINATION: LINKAGE WITH SOMATIC HYPERMUTATION, Tasuku Honjo, Kazuo Kinoshita, and Masamichi Muramatsu

165

INNATE IMMUNE RECOGNITION, Charles A. Janeway, Jr. and Ruslan Medzhitov

KIR: DIVERSE, RAPIDLY EVOLVING RECEPTORS OF INNATE AND ADAPTIVE IMMUNITY, Carlos Vilches and Peter Parham ORIGINS AND FUNCTIONS OF B-1 CELLS WITH NOTES ON THE ROLE OF CD5, Robert Berland and Henry H. Wortis E PROTEIN FUNCTION IN LYMPHOCYTE DEVELOPMENT, Melanie W. Quong, William J. Romanow, and Cornelis Murre

197 217 253 301

LYMPHOCYTE-MEDIATED CYTOTOXICITY, John H. Russell and Timothy J. Ley

SIGNAL TRANSDUCTION MEDIATED BY THE T CELL ANTIGEN x

323

P1: FRK

Annu. Rev. Immunol. 2002.20:795-823. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

February 1, 2002

12:47

Annual Reviews

AR152-FM

CONTENTS

xi

RECEPTOR: THE ROLE OF ADAPTER PROTEINS, Lawrence E. Samelson INTERACTION OF HEAT SHOCK PROTEINS WITH PEPTIDES AND ANTIGEN PRESENTING CELLS: CHAPERONING OF THE INNATE AND ADAPTIVE IMMUNE RESPONSES, Pramod Srivastava CHROMATIN STRUCTURE AND GENE REGULATION IN THE IMMUNE SYSTEM, Stephen T. Smale and Amanda G. Fisher PRODUCING NATURE’S GENE-CHIPS: THE GENERATION OF PEPTIDES FOR DISPLAY BY MHC CLASS I MOLECULES, Nilabh Shastri, Susan

371

Schwab, and Thomas Serwold

395 427

463

THE IMMUNOLOGY OF MUCOSAL MODELS OF INFLAMMATION, Warren Strober, Ivan J. Fuss, and Richard S. Blumberg T CELL MEMORY, Jonathan Sprent and Charles D. Surh

GENETIC DISSECTION OF IMMUNITY TO MYCOBACTERIA: THE HUMAN MODEL, Jean-Laurent Casanova and Laurent Abel ANTIGEN PRESENTATION AND T CELL STIMULATION BY DENDRITIC CELLS, Pierre Guermonprez, Jenny Valladeau, Laurence Zitvogel, Clotilde Th´ery, and Sebastian Amigorena

495 551 581

621

NEGATIVE REGULATION OF IMMUNORECEPTOR SIGNALING, Andr´e Veillette, Sylvain Latour, and Dominique Davidson

669

CPG MOTIFS IN BACTERIAL DNA AND THEIR IMMUNE EFFECTS, Arthur M. Krieg

PROTEIN KINASE Cθ

709 IN

T CELL ACTIVATION, Noah Isakov and Amnon

Altman

RANK-L AND RANK: T CELLS, BONE LOSS, AND MAMMALIAN EVOLUTION, Lars E. Theill, William J. Boyle, and Josef M. Penninger PHAGOCYTOSIS OF MICROBES: COMPLEXITY IN ACTION, David M. Underhill and Adrian Ozinsky

761 795 825

STRUCTURE AND FUNCTION OF NATURAL KILLER CELL RECEPTORS: MULTIPLE MOLECULAR SOLUTIONS TO SELF, NONSELF DISCRIMINATION, Kannan Natarajan, Nazzareno Dimasi, Jian Wang, Roy A. Mariuzza, and David H. Margulies

853

INDEXES Subject Index Cumulative Index of Contributing Authors, Volumes 1–20 Cumulative Index of Chapter Titles, Volumes 1–20

ERRATA An online log of corrections to Annual Review of Immunology chapters (if any, 1997 to the present) may be found at http://immunol.annualreviews.org/

887 915 925

8 Feb 2002

10:13

AR

AR152-25.tex

AR152-25.SGM

LaTeX2e(2001/05/10) P1: IKH 10.1146/annurev.immunol.20.103001.114744

Annu. Rev. Immunol. 2002. 20:825–52 DOI: 10.1146/annurev.immunol.20.103001.114744 c 2002 by Annual Reviews. All rights reserved Copyright °

PHAGOCYTOSIS OF MICROBES: Complexity in Action

Annu. Rev. Immunol. 2002.20:825-852. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

David M. Underhill∗ and Adrian Ozinsky Institute for Systems Biology, 1441 North 34 Street, Seattle, Washington 98103; e-mail: [email protected], [email protected]

Key Words innate immunity, macrophage, inflammation, phagocytes, pathogens ■ Abstract The phagocytic response of innate immune cells such as macrophages is defined by the activation of complex signaling networks that are stimulated by microbial contact. Many individual proteins have been demonstrated to participate in phagocytosis, and the application of high-throughput tools has indicated that many more remain to be described. In this review, we examine this complexity and describe how during recognition, multiple receptors are simultaneously engaged to mediate internalization, activate microbial killing, and induce the production of inflammatory cytokines and chemokines. Many signaling molecules perform multiple functions during phagocytosis, and these molecules are likely to be key regulators of the process. Indeed, pathogenic microorganisms target many of these molecules in their attempts to evade destruction.

INTRODUCTION Phagocytosis as a mechanism of innate immune defense has been appreciated since the late nineteenth century when Eli Metchnikov first proposed that mobile phagocytic cells survey tissues for foreign particles and engage in pitched battles with potential pathogens (1). Since then, phagocytosis has served as the classic model of microbe-innate immune interactions, and enormous progress has been made toward understanding the consequences of this interaction. Phagocyte-microbe contact is accompanied by intracellular signals that trigger cellular processes as diverse as cytoskeletal rearrangement, alterations in membrane trafficking, activation of microbial killing mechanisms, production of pro- and anti-inflammatory cytokines and chemokines, activation of apoptosis, and production of molecules required for efficient antigen presentation to the adaptive immune system (Figure 1) (2, 3). Even the fundamental process of internalizing particles proceeds through a variety of distinct molecular and morphological processes (2). Internalization of IgG-opsonized particles is characterized by phagocyte membrane extension ∗

To whom correspondence should be addressed.

0732-0582/02/0407-0825$14.00

825

1 Feb 2002

10:3

Annu. Rev. Immunol. 2002.20:825-852. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

826

AR

AR152-25.tex

UNDERHILL

¥

AR152-25.SGM

LaTeX2e(2001/05/10)

P1: IKH

OZINSKY

Figure 1 Receptor and signaling interactions during phagocytosis of microbes. Multiple receptors simultaneously recognize microbes both through direct binding and by binding to opsonins on the microbe surface. Receptor engagement induces many intracellular signals, and several molecules are utilized in many pathways. Signaling during phagocytosis may subsequently serve to activate or inhibit further phagocytosis and microbe-induced responses. Many pathogenic microbes actively regulate phagocyte responses.

around the particles, a requirement for Syk tyrosine kinase, and by production of pro-inflammatory mediators. Conversely, phagocytosis of complement-opsonized particles occurs without appreciable membrane extension (the particle appears to sink into the cell), does not require Syk, and often is not accompanied by the production of inflammatory mediators (2). The diversity of phagocytic mechanisms presents a challenge to those who strive to elucidate the underlying principles of the process. Roles for many receptors and many signaling molecules have been described, and key signaling molecules are emerging as regulators of multiple phagocytic responses. Recent applications of high-throughput analytical tools and a willingness to incorporate complexity into analysis of the process are accelerating discovery in the field. In one recent study, Desjardins and coworkers purified macrophage phagosomes containing latex beads and identified more than 140 proteins associated with the phagosome by two-dimensional electrophoresis and mass spectrometry

1 Feb 2002

10:3

AR

AR152-25.tex

AR152-25.SGM

LaTeX2e(2001/05/10)

P1: IKH

Annu. Rev. Immunol. 2002.20:825-852. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

PHAGOCYTOSIS

827

(4). These investigators identified many proteins not previously known to be associated with phagosomes, as well as many novel proteins (4). Further characterization of one protein, flotillin-1, which is known to be enriched in lipid rafts, revealed that lipid rafts accumulate in phagosomes as part of the maturation process and that other proteins such as V-ATPase subunits, which also are part of the maturation process, are also associated with lipid rafts (5). In a related approach, Morrissette et al. generated a large panel of monoclonal antibodies to purified phagosomes (6). Subsequent analysis of one of the antibody targets revealed a novel variant of amphiphysin (amphiphysin IIm) that, by recruiting dynamin 2 to membranes, is required for internalization of particles (7, 8). Genetically exploitable model organisms including Dictyostelium, Drosophila, and C. elegans are phagocytic (or have phagocytic cells), and screens in these organisms have identified novel genes involved in phagocytosis of bacteria and apoptotic cells (9–12). Thus, Cornillon et al. recently identified a membrane protein in Dictyostelium called PHG1 (which has many human homologues of unknown function) that is required for adhesion to particles during phagocytosis. Another broad-based technology, cDNA array expression analysis, is increasingly used to analyze the consequences of microbe-phagocyte interactions; these studies expand our understanding of the complexities of phagocyte inflammatory responses (13–17). In one study, Ehrt et al. identified over 200 genes with altered expression during phagocytosis of latex particles by mouse bone marrow–derived macrophages, and over 600 genes altered when cells were exposed to Mycobacterium tuberculosis (17). Pathogenic microbes have evolved many mechanisms for reducing the efficiency of phagocytosis or effective killing, and these microbes are also useful tools for discovering molecules important for regulating immune responses. During microbial contact, many parallel signaling pathways are simultaneously activated that together define the phagocyte response and regulate internalization (Figure 1). This review does not aspire to be a comprehensive analysis of all the complexities associated with phagocytosis. Instead, we present phagocytosis as a complex system with special emphasis on four sources of that complexity. First, many different receptors recognize microbes, and phagocytosis is usually mediated simultaneously by multiple receptors. Second, different microbe-recognition receptors induce different signaling pathways, and these signals interact cooperatively (and sometimes destructively) to mediate ultimate responses to particles. Third, microbe recognition is coupled (either directly through phagocytic receptors or indirectly through coreceptors) to inflammatory responses that in turn affect the efficiency of particle internalization by the phagocyte or neighboring phagocytes. Fourth, many pathogenic microbes actively attempt to regulate the mechanisms of phagocytosis to evade destruction. In this review we have focused our discussion specifically on phagocytosis of microbes. Phagocytosis also is required for normal clearance of apoptotic cells, a process in which many of the same levels of complexity apply. For a proper discussion of the phagocytosis of apoptotic cells, the reader is referred to several recent reviews (2, 18, 19).

1 Feb 2002

10:3

AR

828

AR152-25.tex

UNDERHILL

¥

AR152-25.SGM

LaTeX2e(2001/05/10)

P1: IKH

OZINSKY

Annu. Rev. Immunol. 2002.20:825-852. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

MICROBIAL CONTACT Phagocytes express a broad spectrum of receptors that participate in particle recognition and internalization (Table 1). Some of these receptors are capable of transmitting intracellular signals that trigger phagocytosis, while other receptors appear primarily to participate in binding or to increase the efficiency of internalization. The most reliable method for establishing the phagocytic capacity of a specific receptor is to express it in a nonphagocytic cell and demonstrate that it confers on the cell the ability to internalize specific target particles. While this method has been used to analyze phagocytosis through receptors such as Fc-receptors, complement receptor 3, and the mannose receptor, in many cases the ability of a receptor to function (or not) in this type of assay has not been assessed. Full descriptions of every receptor implicated in phagocytosis are beyond the scope of this review, but we have tried in this section to describe several of the main classes of phagocytic receptors and discuss their capacities to interact with each other.

TABLE 1 Phagocytic receptors for microbes Receptors that participate in phagocytosis of microbes

Ligands

References

Fc-Receptors: Fcγ RI (CD64) Fcγ RII* (CD32) Fcγ RIII (CD16) FcεRI FcεRII (CD23) FcαRI (CD89)

IgG-, CRP-, SAP-opsonized particles IgG-, CRP-, SAP-opsonized particles IgG-, CRP-, SAP-opsonized particles IgE-opsonized particles IgE-opsonized particles IgA-opsonized particles

(20–23) (20–23) (20–23) (24) (25) (26)

Complement receptors: CR1 (CD35) CR3 (α Mβ 2, CD11b/CD18, Mac1) CR4 (α Xβ 2, CD11c/CD18, gp150/95)

MBL-, C1q-, C4b-, C3b-opsonized particles iC3b-opsonized particles iC3b-opsonized particles

(27) (28) (29)

Fibronectin/Vitronectin-opsonized particles

(30)

Bacteria, LPS, Lipoteichoic Acid Bacteria

(31) (32)

Mannose receptor (CD206)

Mannan

(33)

Dectin-1

β1,3-glucan

(34)

CD14

LPS, peptidoglycan,

(35, 36)

C1qR(P)

C1q, MBL, SPA

(37)

Various integrins: α 5β 1 (CD49e/CD29) α 4β 1 (CD49d/CD29) α vβ 3 (CD51/CD61) Scavenger receptors: SRA MARCO

∗

In humans, Fcγ RIIA is an activating phagocytic receptor, and Fcγ RIIB is an inhibitory receptor. Mice express only Fcγ RIIB. CRP, C-reactive protein; SAP, Serum amyloid P.

1 Feb 2002

10:3

AR

AR152-25.tex

AR152-25.SGM

LaTeX2e(2001/05/10)

P1: IKH

PHAGOCYTOSIS

829

Annu. Rev. Immunol. 2002.20:825-852. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

Fcγ -Receptors IgG-opsonized particles are recognized by several surface receptors that bind to the Fc region of IgG (Fcγ Rs) (20, 38). Phagocytes such as macrophages or neutrophils express different combinations of Fcγ Rs; thus, recognition of IgG-opsonized particles occurs simultaneously through several receptors. Fcγ Rs fall into two classes: (a) receptors that contain ITAM motifs in their intracellular domains that recruit kinases and activate phosphorylation cascades, and (b) receptors that contain ITIM motifs that recruit phosphatases that inhibit signaling (20, 38). Activating receptors with high affinity (Fcγ RI) and low affinity (Fcγ RIIA and Fcγ RIIIA) bind IgGopsonized particles and trigger internalization through actin polymerization beneath the particle, membrane recruitment to the site of particle contact, membrane extension outward to surround the particle, and particle engulfment (2). The efficiency of the process is regulated by coligation of the inhibitory Fcγ R (Fcγ RIIB) that recruits the phosphatase SHIP that blocks phophoinositide signaling (20). Thus, relative expression of activating and inhibiting Fcγ Rs determines the threshold for phagocytosis and inflammatory responses to IgG-opsonized particles.

Complement Receptors Complement proteins in the serum can opsonize microbes through antibodydependent and antibody-independent mechanisms; complement-opsonized particles are recognized and internalized via specific complement receptors. Phagocytic complement receptors include complement receptor 1 (CR1) expressed on erythrocytes, B cells, monocytes, neutrophils, eosinophils, and dendritic cells; complement receptor 3 (CR3, α Mβ 2 integrin, CD11b/CD18, or Mac1) found on monocytes, macrophages, neutrophils, granulocytes, dendritic cells, and NK cells; and complement receptor 4 (CR4, α Xβ 2 integrin, CD11c/CD18, or gp150/95), which has not been as well characterized as CR3 (39). CR2 (CD21) has not been described as a phagocytic receptor. CR1 is a single-chain transmembrane molecule with a large extracellular lectin-like ligand recognition domain and a short intracellular tail (27). CR1 binds a broad spectrum of microbial opsonins including complement components C1q, C4b, and C3b, as well as mannan-binding lectin (MBL) (27, 40). CR1 alone is unable to mediate internalization of a particle without additional signals, although CR1 ligation enhances Fc-receptor–mediated phagocytosis, and it can mediate phagocytosis of MBL-opsonized particles in fibronectin-treated PMNs (27). The integrins CR3 and CR4 are both heterodimers consisting of a shared beta chain (β2, CD18) paired with specific alpha chains, α M/CD11b, and α X/CD11c, respectively, and both receptors recognize iC3b. Like CR1, these complement receptors require additional signals in order to mediate internalization of complementopsonized particles. Internalization signaled by CR3 requires a second activation step that increases the number of receptors at the cell surface (30, 41, 42)—the affinity of the receptors (43)—and allows the receptors to trigger phagocytosis

1 Feb 2002

10:3

Annu. Rev. Immunol. 2002.20:825-852. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

830

AR

AR152-25.tex

UNDERHILL

¥

AR152-25.SGM

LaTeX2e(2001/05/10)

P1: IKH

OZINSKY

(44, 45). Inflammatory cytokines (TNFα), microbial products (LPS), and adhesion (fibronectin) stimulate phagocytosis through CR3, demonstrating how heterogeneous cellular processes influence phagocytosis (44, 45). In vitro, this activation signal can be stimulated with phorbol esters and is therefore likely to involve protein kinase C activation (2, 44). Recently, Caron et al. have demonstrated that activation of Rap1, a small molecular weight GTPase in the Ras family, is required for PMA-stimulated phagocytosis of complement-opsonized particles (46). Expression of constitutively active R-ras or Rap1 is sufficient to activate complement receptors for phagocytosis; expression of dominant-negative Rap1 (but not R-ras) blocks PMA-activated complement receptor-mediated phagocytosis (46). In addition to the variety of Fc-receptors and complement receptors that participate separately in particle internalization, it has been observed for more than 25 years that complement receptor and Fc-receptor coligation can produce cooperative effects (47). For example, macrophages do not internalize particles coated with suboptimal concentrations of IgG, but do internalize these particles when they also are coated with complement (47). The experiments suggest that Fc-receptor ligation signals activation of the complement receptors for internalization. Similar data have recently been reported for particles coated with suboptimal concentrations of IgG together with MBL, suggesting that Fc-receptors also cooperate with CR1 (27). Thus, simultaneous binding by different phagocytic receptors produces synergistic effects.

Scavenger Receptors Defined originally by their ability to bind and internalize modified lipoproteins such as acetylated low-density lipoprotein, scavenger receptors additionally bind such diverse ligands as polyribonucleotides, lipopolysaccharide, and silica particles (48). Two members of the scavenger receptor family have been implicated in binding and internalizing microbes. Scavenger receptor A (SR-A) is a transmembrane homo-trimer with an extended extracellular domain composed of a collagenous triple-helix. It is expressed on most macrophages and binds whole bacteria as well as the microbial cell wall components, lipoteichoic acid and LPS (49, 50). Macrophages from mice lacking SR-A are less efficient at phagocytosing heat-killed E. coli (51, 52). MARCO (macrophage receptor with collagenous structure), another member of the class A scavenger receptor family, also participates in the phagocytosis of microbes. MARCO is expressed constitutively on macrophage subpopulations in the marginal zone of the spleen (53) and can be induced in other macrophages by exposure to inflammatory stimuli such as LPS (32). MARCO binds to a variety of particles including Gram-positive bacteria, Gram-negative bacteria, and artificial particles such as latex; inhibitory antibodies to MARCO significantly block internalization of these targets (32, 54). The precise ligand recognized by MARCO has not been defined. Whereas it is clear that scavenger receptors participate in the phagocytosis of microbes, it is likely that they contribute to binding, while coreceptors generate the internalization signals.

1 Feb 2002

10:3

AR

AR152-25.tex

AR152-25.SGM

LaTeX2e(2001/05/10)

P1: IKH

PHAGOCYTOSIS

831

Unlike the mannose receptor or CR3 that reconstitute phagocytosis in nonphagocytic cells, expression of SRA or MARCO confers binding without significant internalization (32, 51, 53).

Annu. Rev. Immunol. 2002.20:825-852. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

Lectins Mammalian phagocytes express a wide variety of surface lectins that mediate detection of self and foreign carbohydrates, and these receptors cooperate in detection of microbes. Zymosan is a Saccharomyces cerevisiae yeast cell wall particle that is made up primarily of α-mannan /mannoproteins and β-glucans (55). Soluble forms of both α-mannan and β-glucan inhibit phagocytosis of zymosan by macrophages, suggesting that receptors for both of these sugars participate in particle recognition and uptake (56–58). The mannose receptor (that binds α-mannan) and dectin-1 (that binds β-glucan) have been demonstrated to mediate phagocytosis of yeast and zymosan (33, 34). The macrophage mannose receptor is a type I transmembrane protein with a short 45 amino acid cytoplasmic tail and an extracellular domain consisting of eight C-type lectin carbohydrate recognition domains, a short amino terminal cysteine rich region, and a fibronectin type II repeat (139, 140). The receptor is expressed on subpopulations of macrophages and dendritic cells. Expression of mannose receptor in normally nonphagocytic COS cells is sufficient to allow internalization of zymosan, and expression of a tailless mutant mannose receptor permits zymosan binding but not internalization (33). Additionally, expression of a chimeric receptor consisting of the extracellular domain of Fcγ RI fused to the transmembrane and intracellular domains of the mannose receptor facilitates uptake of IgG-opsonized particles (59). Together these observations indicate that signals initiated by mannose receptor ligation are sufficient to induce particle internalization. The molecular mechanisms by which mannose receptors activate downstream signals for particle internalization remain to be described. Brown & Gordon have recently defined dectin-1 as a β-glucan binding lectin capable of mediating phagocytosis of zymosan when the receptor is expressed in normally nonphagocytic cells (34). Dectin-1 was originally defined as a dendritic cell–specific receptor with a short extracellular domain consisting of a single CRD and a short cytoplasmic tail containing a putative ITAM (60, 61). The receptor binds T cells in addition to β-glucans (although at distinct sites), and this interaction may deliver costimulatory signals for T cell activation (61, 62). Subsequent analysis determined that dectin-1 is widely expressed on cells of myeloid lineage and may be the predominant β-glucan receptor for phagocytosis (34, 62). Still, there is evidence that other receptors may participate in recognition of β-glucans, suggesting that responses to β-glucans may be regulated by the relative expression of the different receptors. CR3 has a high-affinity (5 × 10−8 M) β-glucan binding site on the α M chain (39) and may participate in recognition of zymosan. Antibodies to CR3 block binding and internalization of unopsonized zymosan (63); patients with a deficiency in β 2 integrins (and thus CR3) show defects in phagocytosis of

1 Feb 2002

10:3

832

AR

AR152-25.tex

UNDERHILL

¥

AR152-25.SGM

LaTeX2e(2001/05/10)

P1: IKH

OZINSKY

unopsonized zymosan (64). It therefore seems likely that these two β-glucan receptors (together with the mannose receptor) coordinate phagocytosis of zymosan in macrophages.

Annu. Rev. Immunol. 2002.20:825-852. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

Other Integrins Both fibronectin and vitronectin can nonspecifically opsonize pathogens and cell debris. Phagocytes recognize and ingest these particles primarily through α 5β 1 and α vβ 3 integrins, and this activity can be reconstituted by expressing the receptors in nonphagocytic cells (30, 65). Particle ingestion requires a second signal that can be provided by activation of protein kinase C with phorbol esters (30, 65). There are complex interactions between α 5β 1 and α vβ 3. When expressed together, α 5β 1 primarily mediates internalization. However, under certain circumstances, ligation of α vβ 3 delivers an inhibitory signal that blocks internalization through α 5β 1 (30, 65). Inhibitory signaling through α vβ 3 requires serine phosphorylation of the β 3 cytoplasmic tail (66) and may, in part, be due to the association of α vβ 3 with CD47 (also called integrin associated protein, IAP), a G-protein coupled receptor implicated in thrombospondin-induced inhibition of inflammatory signaling. Through its ligand SIRPα, CD47 inhibits both Fc- and complement receptor–mediated phagocytosis (67, 68). The decisions to activate inflammatory signals are complex, and α vβ 3 can participate in both inflammatory and anti-inflammatory responses. Ligation of α vβ 3 in PMNs activates Fc-receptor-mediated phagocytosis and production of reactive oxygen intermediates (69, 70). Conversely, α vβ 3 and another vitronectin-binding integrin, α vβ 5, can interact with CD36 (a class B scavenger receptor) to mediate phagocytosis of apoptotic cells, which is accompanied by anti-inflammatory signals (71, 72).

ENGULFMENT Particle internalization is accompanied by activation of many signaling pathways that together orchestrate rearrangement of the actin cytoskeleton, extension of the plasma membrane, and engulfment. For specific discussions of molecules required for these processes, the reader is directed to a number of recent reviews (2, 3, 20, 38). A brief survey of molecules that have been functionally revealed to be required for efficient phagocytosis of various types of particles in just the past two years shows how rapidly our understanding of the complexity of phagocytic signaling is growing (Table 2). Dozens (if not hundreds) of signaling molecules including actin binding proteins, membrane traffic regulators, ion channels, kinases, and lipases are activated during phagocytosis of complex particles (such as opsonized bacteria) and may contribute to efficient internalization. Linking these proteins in a model of phagocytosis is a challenge, and identifying the key regulators is a task that might be best suited to computational modeling. However, certain signaling molecules stand out both as participating in phagocytosis and as participating in many other signaling pathways. Phosphoinositide 3-kinase

1 Feb 2002

10:3

AR

AR152-25.tex

AR152-25.SGM

LaTeX2e(2001/05/10)

P1: IKH

PHAGOCYTOSIS

833

TABLE 2 Proteins recently functionally implicated in phagocytosis of microbes Protein

Annu. Rev. Immunol. 2002.20:825-852. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

Amphiphysin IIm

Reference (7)

Protein

Reference

MLCK

(96)

Arp2/3

(74)

NSF

(73)

BLNK

(76)

p60(SRC)

(75)

CapG

(78)

PAG3

(77)

CD47

(68)

PKCα

(79)

ClassI/ClassII PI 3-kinases

(80)

pp155(FAK)

(75)

Cofilin

(82)

Pyk2

(81)

COP1

(84)

Rab11

(83)

CSK

(86)

Rab5

(85)

Fgr

(88, 89)

Rab7

(87)

Gelsolin

(78)

Rap1

(46)

Hck

(88)

SHIP

(90)

Iba1

(91)

SIRPα

(68)

LAT

(92)

SLP76

(76)

LIM kinase

(82)

VAMP3

(93)

Lyn

(88)

WASP

(94, 95)

(PI 3-kinase), phospholipase C (PLC), Rho GTPases, and PKC are integration points for regulating of phagocytosis. Furthermore, these molecules not only orchestrate the mechanics of particle ingestion, they also regulate inflammatory responses and microbial killing.

Phosphoinositide 3-Kinase and Phospholipase C PI 3-kinase catalyzes phosphorylation of PI(4,5)P2 to PI(3,4,5)P3, a phospholipid important in recruiting signaling molecules such as the kinase AKT/PKB to specific regions of membranes (97). Pharmacological inhibition of PI 3-kinase blocks phagocytosis of IgG- and complement-opsonized particles, unopsonized zymosan, and bacteria, although the requirement may not be absolute because the uptake of IgG-opsonized particles smaller than 3 microns across is less affected than the uptake of larger particles (8, 98, 99). PI 3-kinase is not required for particle binding or initial actin polymerization beneath the particle, suggesting that initial phagocytic signaling is intact. Instead, inhibition of PI 3-kinase blocks membrane extension and fusion behind the bound particle, perhaps due to a failure to insert new membrane at the site of particle internalization (8, 98, 99). In recent work using cells from knockout mice and microinjection of inhibitory antibodies, Vieira and colleagues have demonstrated that these stages of phagocytosis require class

1 Feb 2002

10:3

Annu. Rev. Immunol. 2002.20:825-852. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

834

AR

AR152-25.tex

UNDERHILL

¥

AR152-25.SGM

LaTeX2e(2001/05/10)

P1: IKH

OZINSKY

I PI 3-kinase, while a second type of PI 3-kinase, class II/VPS34, is required for phagosome maturation (80). Phosphoinositide-specific phospholipase C (PI-PLC) mediates cleavage of PI (4, 5)P2 resulting in release of IP3 and DAG, second messengers that mobilize intracellular Ca++ stores and activate protein kinase C family members, respectively. PI-PLC is recruited to phagosomes containing IgG-opsonized particles, and inhibition of its activity blocks particle internalization (100). Like protein kinase C inhibitors, PLC inhibitors completely block the formation of actin filaments beneath the site of particle contact, suggesting that the main role of PLC in particle internalization is to activate protein kinase C (100). In addition to being required for the mechanical aspects of particle internalization, PI 3-kinase and PI-PLC also are implicated in pro-inflammatory signaling induced by particulate stimuli. Thus, PI 3-kinase is recruited to Toll-like receptors when cells are stimulated with heat-killed S. aureus, and activation of PI 3-kinase has been implicated in NF-κB-mediated cytokine production (101). Similarly, PI-PLC activity is required for microbe-induced pro-inflammatory signaling in macrophages primarily due to its role in activating PKC (102, 103).

Rho GTPases Many of the signaling molecules that mediate phagocytosis have well-defined roles in other cellular signaling processes, making it likely that other ongoing cellular processes such as cell division, signaling by hormones, adhesion, or active secretion might affect phagocytosis by pre-activating or sequestering specific molecules. Members of the Rho family of small molecular weight GTPases are key regulators of the actin cytoskeleton in adhesion, membrane ruffling, and stress fiber formation, and Rho family GTPases cooperate with Ras family GTPases in the regulation of gene expression and proliferation (104). The Rho family/GTPases (Cdc42, Rac, and Rho) play central roles in phagocytosis, and due to their interactions with other signaling pathways, are likely to be points of intersection for signaling pathways that regulate phagocytic efficiency (104, 105). Fc-receptormediated phagocytosis is blocked by expression of inhibitory mutants of Cdc42 and Rac1 in macrophages and mast cells (106–109). Intriguingly, Rho is not required for Fc-receptor-mediated phagocytosis because expression of C3 transferase (a Rho-specific inhibitor) does not block internalization of IgG-opsonized particles (106). Conversely, internalization of complement-opsonized particles is regulated by Rho but does not require active Cdc42 or Rac (106). Thus, while Rho GTPases are general regulators of phagocytosis, regulation of individual family members permits signals from different phagocytic receptors to be controlled independently. Rho GTPases can be regulated at several levels. Like other small molecular weight GTPases, Rho proteins cycle between an inactive, GDP-bound state and an active GTP-bound state. Activation is stimulated by GDP-GTP exchange factors (GEFs), while GTPase-activating proteins (GAPs) inactivate the proteins (105). Indeed, the pathogen Yersinia pseudotuberculosis injects a Rho family GAP, YopE,

1 Feb 2002

10:3

AR

AR152-25.tex

AR152-25.SGM

LaTeX2e(2001/05/10)

P1: IKH

PHAGOCYTOSIS

835

into the cytoplasm of macrophages to inhibit its phagocytosis (110). Rho GEFs and GAPs are regulated by many signaling pathways including growth factor receptor signaling, adhesion-dependent signaling, and cell cycle progression, providing intersections at which various internal and external stimuli may regulate phagocytosis (104). Similarly, Rho GTPase-induced activation of downstream signals may be affected by other coincident signals since, for example, both Rho GTPases and Ras GTPases activate MAP kinase signaling (104).

Annu. Rev. Immunol. 2002.20:825-852. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

Protein Kinase C Protein kinase C family members are required for phagocytosis, but also participate in many other signal transduction pathways. General inhibitors of PKC activity inhibit internalization of IgG-opsonized, and complement-opsonized particles as well as unopsonized zymosan particles (111, 112). PKC is required at the earliest stages of particle internalization since inhibition of PKC blocks the formation of actin filaments beneath the site of particle binding (112). The precise isoform(s) of PKC required for internalization is not yet entirely clear. Of the 12 isoforms of protein kinase C (PKC) described, at least 5 (PKCα, -β, -ε, -δ, and -ζ ) are expressed in macrophages and are recruited to membranes during phagocytosis (112–114). Expression of a dominant negative form of PKCα inhibits ingestion of IgG-opsonized particles (79), but Larsen et al. note that IgG-opsonized particle uptake is Ca++-independent, suggesting a role for the novel (Ca++-independent) PKCδ and PKCε (115). The PKC isoforms required may depend on the differentiation state of the macrophage, since Melendez and coworkers have observed that Fc-receptor ligation activates different PKC isoforms during different stages of monocyte-macrophage differentiation (113). PKCs also are required for cytokine production and activation of killing mechanisms induced during phagocytosis of microbes. Pharmacological inhibitors of PKC block both LPS-induced production of COX-2, TNF-α, and IL-1β and activation of the respiratory burst (116–118). The role for PKCα was shown using expression of dominant-negative PKCα in macrophages to inhibit LPS- and Fc-receptor-induced cytokine production (118– 120). PKCs are key participants in numerous signaling pathways to the actin cytoskeleton and the nucleus, including signals stimulated by hormones, cytokines, and adhesion, suggesting multiple levels of regulation of phagocytic efficiency (121, 122). Indeed, as discussed above, complement receptors are not phagocytic unless a second, PKC-dependent signal is provided that may be delivered by cytokines or adhesion to fibronectin (44, 45).

COUPLING INFLAMMATION TO PHAGOCYTOSIS Microbe internalization by phagocytes is usually accompanied by the production of pro-inflammatory signals and activation of antimicrobial mechanisms. Certain phagocytic receptors such as Fc-receptors trigger inflammatory responses directly (38, 123), whereas others such as complement receptors often do not stimulate

1 Feb 2002

10:3

836

AR

AR152-25.tex

UNDERHILL

¥

AR152-25.SGM

LaTeX2e(2001/05/10)

P1: IKH

OZINSKY

inflammatory responses. In many cases, the decisions to activate inflammatory responses during phagocytosis are regulated by additional receptors (that are not themselves phagocytic), such as Toll-like receptors (TLRs), that are specifically recruited to phagosomes (124, 125). Phagocytic receptors often function cooperatively to evoke inflammatory responses, and it is interesting to note that patients deficient in β 2 integrins (leukocyte adhesion deficiency) also have reduced Fc-receptor-mediated inflammatory responses (126–128).

Annu. Rev. Immunol. 2002.20:825-852. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

Superoxide Production Professional phagocytes including monocytes, macrophages, and especially neutrophils kill newly internalized pathogens in part by the production of caustic reactive superoxide ions produced by assembly of the NADPH oxidase on phagosomal membranes (129, 130). In unstimulated cells, components of the NADPH oxidase are distributed in the membrane (gp91phox and p22phox) and in the cytosol (p67phox, p40phox, p47phox, and Rac2). Stimulus-induced phosphorylation of p47phox induces translocation of a trimeric p67phox, p40phox, p47phox cytosolic complex to the membrane components (130). Several kinases including PKC, PKA, p21-activated kinase (PAK), and PI 3-kinase-stimulated kinases can phosphorylate p47phox and are required for activation of the complex (130). Rac2, activated by a GDP-GTP exchange factor, also translocates to the membrane complex where it is required for electron-transfer in the active complex (131). Besides the phox proteins, all of these proteins also are important regulators of the actin cytoskeleton, and nearly all stimuli that activate the oxidase also induce profound alterations in the actin cytoskeleton (130). In some cases, disruption of the actin cytoskeleton blocks activation of the oxidase by particulate stimuli (132, 133). The link to the actin cytoskeleton may be the observation that in the cytosol, p40phox and p67phox are associated with coronin, another protein that plays an important role in regulation of the actin cytoskeleton and that strongly localizes to forming phagosomes (6, 134). Phagosome formation is not a prerequisite for superoxide production, because the NADPH oxidase may also be activated by soluble stimuli including PMA and fMLP (130). Phagocytic receptors such as Fc-receptors activate the NADPH oxidase in macrophages and neutrophils (127, 135). However, phagocytosis is not always accompanied by NADPH oxidase activation because internalization though the CR3 in macrophages does not lead to the production of superoxide (136, 137). However, the effects of CR3 ligation are context-specific, and in neutrophils complementopsonized particles activate the NADPH oxidase (133, 138). It is less clear whether other phagocytic receptors activate the NADPH oxidase. It has been reported that ligation of the mannose receptor is sufficient to induce a respiratory burst, but these studies have depended on stimulation of macrophages with complex particles (such as zymosan) that engage multiple receptors (139, 140). A recent report by Maridonneau-Parini and coworkers suggests that mannose receptor-mediated phagocytosis does not induce microbicidal mechanisms (141). They observed that

1 Feb 2002

10:3

AR

AR152-25.tex

AR152-25.SGM

LaTeX2e(2001/05/10)

P1: IKH

PHAGOCYTOSIS

837

Annu. Rev. Immunol. 2002.20:825-852. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

internalization of trimannoside-coated latex beads by human peripheral bloodderived macrophages (a process that was fully blocked by soluble mannan and thus was mediated by a mannose receptor) was not accompanied by superoxide production (141). Zymosan internalization was partially blocked by either soluble mannan or the soluble β-glucan (laminarin) and thus occurred through cooperation between a mannose receptor and a β-glucan receptor. However, zymosan-induced oxidase activation was inhibited only by laminarin, suggesting that a β-glucan receptor activates the oxidase.

Cytokine and Chemokine Production Activation of gene transcription and the production of cytokines and chemokines during phagocytosis is a critical feature in the development of an effective immune response. While some phagoctytic receptors themselves may trigger cytokine and chemokine production, others require additional coreceptors. Toll-like receptors (TLRs) are a family of innate immune recognition receptors that are required for detection of a broad range of microbial products including lipopolysaccharide, peptidoglycan, and bacterial lipopeptides (142, 143), and several TLR family members are actively recruited to phagosomes during microbe internalization where they sample the contents of the phagosomes to determine the nature of the microbes being ingested (Figure 2) (124, 125). Recruitment of TLRs to phagosomes provides a mechanism by which phagocytosis and associated inflammatory responses can be linked, although recruitment does not require TLR activation. TLRs also are recruited to phagosomes during internalization of IgG-opsonized sheep erythrocytes by Fc-receptors, a process that does not stimulate TLRs, suggesting that TLR recruitment is a general feature of phagocytosis and that TLRs are simply poised to recognize ligands should they be present in the phagosome (125). In the case of phagocytosis of IgG-oponsized microbes, it is likely that both Fc-receptors and TLRs are engaged on individual phagosomes, although the potential for synergistic inflammatory signaling by these receptors has not been examined directly. Nonetheless, it is clear that internalization is not required for some inflammatory signaling. In macrophages, inhibition of particle internalization with the PI 3-kinase inhibitor, wortmannin, or by expression of dominant negative dynamin does not block TNF-α production induced by zymosan (8). Conversely, inhibition of TLR2, a receptor required for inflammatory responses to zymosan, does not block phagocytosis of zymosan (124). Despite the ability to dissociate particle internalization and inflammatory signaling, these two processes share spatial, temporal, and functional relationships, as well as many common signaling molecules. This suggests that signaling by phagocytic receptors may influence or modify pro-inflammatory signaling through other receptors. One example of this is the functional interaction between TLR4 (a receptor required for inflammatory responses induced by LPS) and CR3. Vogel and coworkers have demonstrated that macrophages from mice lacking β 2 integrins (and therefore CR3) respond normally to LPS with respect to induction of some

1 Feb 2002

10:3

AR152-25.tex

UNDERHILL

¥

AR152-25.SGM

LaTeX2e(2001/05/10)

P1: IKH

OZINSKY

Annu. Rev. Immunol. 2002.20:825-852. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

838

AR

Figure 2 Toll-like receptor recruitment to phagosomes during recognition of yeast particles. (a) Epitope-tagged Toll-like receptor 2 (TLR2) expressed in macrophages localizes to the plasma membrane. (b and c) During phagocytosis of zymosan particles (numbered ) TLR2 is recruited to sites of particle contact and is highly enriched on phagosomes. Reprinted by permission: Underhill et al., Nature 1999, 410:811–15.

1 Feb 2002

10:3

AR

AR152-25.tex

AR152-25.SGM

LaTeX2e(2001/05/10)

P1: IKH

Annu. Rev. Immunol. 2002.20:825-852. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

PHAGOCYTOSIS

839

genes such as TNF-α and IP-10, but are deficient in LPS-induced activation of genes such as COX-2, IL-12 p40, and IL-12 p35 (144). Thus, the two receptors cooperate in generating the ultimate inflammatory response. Another example of cooperation between phagocytic and inflammatory signaling receptors is the detection of zymosan. Even though at least two lectin receptors (the mannose receptor and dectin-1) are important for internalization of zymosan, we have demonstrated that induction of cytokines and chemokines in macrophages by zymosan is mediated by cooperative interaction between TLR2 and TLR6 (124, 125). Thus, expression of dominant negative TLR2 or TLR6 in macrophages blocks zymosan-induced activation of NF-κB and production of TNFα without blocking internalization (124, 125). It is likely that various recognition receptors may cooperate in the ultimate generation of pro-inflammatory responses, and it is not yet clear how TLRs and lectin receptors interact. In an added level of complexity, Puzo and coworkers have suggested that mannose receptor ligation may specifically negatively regulate TLR signaling (145). They observed that exposure of dendritic cells to α-mannan or an antibody to the mannose receptor blocked subsequent induction of IL-12 by LPS.

MICROBIAL EVASION OF PHAGOCYTIC MECHANISMS Many pathogenic microbes have developed particular strategies to subvert detection by roving phagocytes that survey tissues for the presence of infection. The diversity of subversion mechanisms utilized by different species of bacteria (and parasites) is testament to the complexity of the phagocytic process. Each encounter has the potential to be different, based on the type of microbe and the many factors that modulate the activation state of the phagocyte. Different bacteria use different strategies to subvert defenses in order to establish their particular niche for proliferation. Some pathogens inhibit phagocytosis altogether, while others orchestrate their own internalization. Other pathogens actively modify the maturation of phagosomes and the inflammatory response to avoid killing. For example, while bacteria such as enteropathogenic E. coli suppress engulfment by inhibiting PI 3-kinase signaling in phagocytes (146), and Yersinia species do so by inhibiting actin polymerization (147), microbes such as Salmonella embrace phagocytosis and survive by warding off destruction within the phagocytic vacuole (148). Yersinia uses a type III secretion system to inject a variety of effectors into macrophages. As mentioned above, YopE is an effector that has Rho-GAP activity and inhibits the Rho family GTPases Rac, Rho, and CDC42 to disrupt the actin cytoskeletal rearrangement needed for phagocytosis (110). In addition, the effector YopH is a phosphatase that inhibits phagocytosis mediated by Fc receptors (149) and complement receptors (150), and it also inhibits the phosphorylation of focal adhesion components implicated in phagocytic signaling, such as focal adhesion kinase and paxillin (151–153). The Salmonella pathogenicity island SPI2 encodes a type III secretion system activated by acidification of the phagocytic vacuole and required for bacterial survival in macrophage phagosomes (154–156). The SPI2 system permits Salmonella to avoid NADPH oxidase-dependent killing by

1 Feb 2002

10:3

Annu. Rev. Immunol. 2002.20:825-852. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

840

AR

AR152-25.tex

UNDERHILL

¥

AR152-25.SGM

LaTeX2e(2001/05/10)

P1: IKH

OZINSKY

interfering with delivery of the oxidase to the vacuole (157). The SPI2 effector SpiC interferes with vesicular trafficking to the Salmonella vacuole and limits vacuole maturation (158). The molecular targets of these effectors, as well as the roles of an additional seven SPI2 effectors (identified by virtue of a consensus signal sequence), remain to be defined (159). Whereas acidification of the vacuole is required for Salmonella to activate SPI2-dependent activity for survival, and also is required for Listeria to activate listeriolysin-mediated disruption of the vacuolar membrane that leads to escape from the vacuole into the cytosol, other bacteria strive to limit the normal decline in vacuolar pH that follows internalization and vacuolar maturation. For Mycobacterium tuberculosis, the inhibition of acidification is an active process that occurs only with live bacteria and correlates with depletion of proton ATPase from the vacuolar membrane. Mycobacteria orchestrate a complex set of changes that make the mycobacterial vacuole permissive for growth (160). The mycobacterial vacuole was thought to be poorly fusogenic due to it not being accessible to markers delivered to lysosomes. However, the mycobacterial vacuole is not completely isolated because it is accessible to fluid phase markers delivered from the surface and to recycling surface membrane proteins such as transferrin receptors and MHC class II molecules. The vacuole also communicates with the secretory pathway and is able to acquire newly synthesized lysosomal constituents, such as cathepsin D, from the biosynthetic pathway (161). The molecular mechanisms that permit mycobacteria to exert these effects have not been precisely defined. One clue comes from the observation that Rab5, a molecule required for endosomal fusion, is maintained on the vacuoles of live mycobacteria and not dead mycobacteria (162, 163). Similarly, the actin-binding protein, coronin-1 (TACO), is retained on mycobacterial vacuoles throughout infection, whereas it normally associates transiently with phagosomes (164). Live mycobacteria also are able to suppress transient Ca++ fluxes (that accompany phagocytosis) and exclude the calcium binding protein, calmodulin, and calmodulin-dependent protein kinase II from the vacuole, whereas dead (or opsonised) mycobacteria do not (165, 166). Legionella, Coxiella, Chlamydia, and Leishmania are other bacteria that also interfere with the maturation of their vacuoles. While we have much to learn both about the mechanisms that underlie these changes and the consequences of their effects, it is clear that many microbes modulate the set of molecules that associate with their vacuoles. Macrophages utilize phagocytic receptors and other pattern recognition receptors, such as TLRs, to identify microbes and trigger immune defenses. Organisms that cause prolonged infections within macrophages must be able to subvert these defenses. Infection with Leishmania parasites is characterized by a markedly suppressed inflammatory response (167, 168). When Leishmania enters macrophages, TLR pathways activate the IL-1α promoter, but the full repertoire of immune responses is suppressed, and the IL-6 promoter and an NF-κB reporter are not induced (T. R. Hawn, personal communication). Furthermore, despite robust promoter activity, the IL-1α cytokine is not secreted from the cell, indicating that Leishmania also interferes with posttranscriptional events regulating cytokine production. By

1 Feb 2002

10:3

AR

AR152-25.tex

AR152-25.SGM

LaTeX2e(2001/05/10)

P1: IKH

Annu. Rev. Immunol. 2002.20:825-852. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

PHAGOCYTOSIS

841

contrast, there is no evidence yet of suppression of TLR signaling during mycobacterial infection. Indeed, TLR pathways are vigorously activated by mycobacterial exposure (169, 170), but somehow the microbe is able to survive the consequences of the wide spectrum of cytokines produced. The Yersinia type III effector, YopJ, is a cysteine protease that blocks ubiquitin-like posttranslational modifications required for regulating pro-inflammatory signaling pathways and thereby blocks the induction of cytokine production (171). YopJ binds and inhibits modification of IκB that regulates NFκB signaling and also inhibits signaling through members of the MAPK kinase family that also are required for the transcriptional activation of many cytokine promoters (172). For their part, immune cells limit evasion by utilizing redundant pathways to coordinate microbial destruction. One important mechanism of microbial killing is the production of reactive nitrogen species activated by a variety of microbial components through TLR pathways (173–175). However, TLR-mediated induction of calcium-independent nitric oxide synthase (iNOS) is not always necessary for killing. In human macrophages, TLR2 activates other potent microbicidal mechanisms that are able to kill mycobacteria (176). Microbes may induce iNOS through TLR2/TLR4/MyD88-independent mechanisms, suggesting the involvement of receptors other than TLRs (177). Such crosstalk between TLRs and other innate immune recognition receptors has long been recognized in killing caused by production of reactive oxygen species, where LPS primes activation of phagocyte oxidase induced by phorbol ester or by phagocytic particles that do not utilize TLR pathways (178).

CONCLUDING REMARKS Phagocytosis is an inherently complex process that requires coordinated activation of signaling leading to events as diverse as actin remodeling, alterations in membrane trafficking, particle engulfment, microbial killing, and production of appropriate inflammatory mediators that direct the adaptive immune response. The consequences of phagocytosis vary, and they depend on the identity of the microbial target and the many factors that modulate the activation state of the phagocyte. Many proteins have been identified that play important roles during phagocytosis, and the application of high-throughput technologies is accelerating their discovery. A central challenge will be to integrate these molecules into pathways and networks that account for the diversity of phagocytic responses. Pathogenic microorganisms have sought out vulnerabilities in phagocytic mechanisms in their attempts to subvert defenses, and the use of overlapping and redundant mechanisms during phagocytosis may reflect evolutionary pressures for the development of a complex, though highly robust system. ACKNOWLEDGMENTS We would like to thank Dr. Alan Aderem for support and comments, and Drs. Kelly Smith and Tom Hawn for critical reading of the manuscript. This work was supported by NIH grants GM62995 and AI25032.

1 Feb 2002

10:3

842

AR

AR152-25.tex

UNDERHILL

¥

AR152-25.SGM

LaTeX2e(2001/05/10)

P1: IKH

OZINSKY

Visit the Annual Reviews home page at www.annualreviews.org

Annu. Rev. Immunol. 2002.20:825-852. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

LITERATURE CITED 1. Stossel TP. 1999. The early history of phagocytosis. In Phagocytosis: The Host, ed. S Gordon, pp. 3–18. Stamford, CT: JAI 2. Aderem A, Underhill DM. 1999. Mechanisms of phagocytosis in macrophages. Annu. Rev. Immunol. 17:593–623 3. Greenberg S. 1999. Modular components of phagocytosis. J. Leuk. Biol. 66:712–17 4. Garin J, Diez R, Kieffer S, Dermine JF, Duclos S, Gagnon E, Sadoul R, Rondeau C, Desjardins M. 2001. The phagosome proteome: insight into phagosome functions. J. Cell Biol. 152:165–80 5. Dermine JF, Duclos S, Garin J, St-Louis F, Rea S, Parton RG, Desjardins M. 2001. Flotillin-1-enriched lipid raft domains accumulate on maturing phagosomes. J. Biol. Chem. 276:18,507–12 6. Morrissette NS, Gold ES, Guo J, Hamerman JA, Ozinsky A, Bedian V, Aderem AA. 1999. Isolation and characterization of monoclonal antibodies directed against novel components of macrophage phagosomes. J. Cell. Sci. 112:4705–13 7. Gold ES, Morrissette NS, Underhill DM, Guo J, Bassetti M, Aderem A. 2000. Amphiphysin IIm, a novel amphiphysin II isoform, is required for macrophage phagocytosis. Immunity 12:285–92 8. Gold ES, Underhill DM, Morrissette NS, Guo J, McNiven MA, Aderem A. 1999. Dynamin 2 is required for phagocytosis in macrophages. J. Exp. Med. 190:1849–56 9. Cornillon S, Pech E, Benghezal M, Ravanel K, Gaynor E, Letourneur F, Bruckert F, Cosson P. 2000. Phg1p is a nine-transmembrane protein superfamily member involved in dictyostelium adhesion and phagocytosis. J. Biol. Chem. 275:34287– 92 10. Reddien PW, Cameron S, Horvitz HR. 2001. Phagocytosis promotes programmed cell death in C. elegans. Nature 412: 198–202

11. Hoeppner DJ, Hengartner MO, Schnabel R. 2001. Engulfment genes cooperate with ced-3 to promote cell death in Caenorhabditis elegans. Nature 412:202– 6 12. Franc NC, Heitzler P, Ezekowitz RA, White K. 1999. Requirement for croquemort in phagocytosis of apoptotic cells in Drosophila. Science 284:1991–94 13. Baltathakis I, Alcantara O, Boldt DH. 2001. Expression of different NF-kappaB pathway genes in dendritic cells (DCs) or macrophages assessed by gene expression profiling. J. Cell Biochem. 83:281–90 14. Wang ZM, Liu C, Dziarski R. 2000. Chemokines are the main proinflammatory mediators in human monocytes activated by Staphylococcus aureus, peptidoglycan, and endotoxin. J. Biol. Chem. 275:20260–67 15. Buates S, Matlashewski G. 2001. General suppression of macrophage gene expression during Leishmania donovani infection. J. Immunol. 166:3416–22 16. Detweiler CS, Cunanan DB, Falkow S. 2001. Host microarray analysis reveals a role for the Salmonella response regulator phoP in human macrophage cell death. Proc. Natl. Acad. Sci. USA 98:5850– 55 17. Ehrt S, Schnappinger D, Bekiranov S, Drenkow J, Shi G, Gingeras T, Gaasterland T, Schoolnik G, Nathan C. 2001. Reprogramming of the macrophage transcriptome in response to interferon-γ and Mycobacterium tuberculosis: signaling roles of nitric oxide synthase-2 and phagocyte oxidase. J. Exp. Med. 194:1123–39 18. Henson PM, Bratton DL, Fadok VA. 2001. The phosphatidylserine receptor: a crucial molecular switch? Nat. Rev. Mol. Cell Biol. 2:627–33 19. Gregory CD. 2000. CD14-dependent clearance of apoptotic cells: relevance to

1 Feb 2002

10:3

AR

AR152-25.tex

AR152-25.SGM

LaTeX2e(2001/05/10)

P1: IKH

PHAGOCYTOSIS

20.

Annu. Rev. Immunol. 2002.20:825-852. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

21.

22.

23.

24.

25.

26.

27.

28.

29.

the immune system. Curr. Opin. Immunol. 12:27–34 Ravetch JV, Bolland S. 2001. IgG Fc receptors. Annu. Rev. Immunol. 19:275– 90 Bharadwaj D, Stein MP, Volzer M, Mold C, Du Clos TW. 1999. The major receptor for C-reactive protein on leukocytes is fcgamma receptor II. J. Exp. Med. 190:585–90 Bharadwaj D, Mold C, Markham E, Du Clos TW. 2001. Serum amyloid P component binds to Fc gamma receptors and opsonizes particles for phagocytosis. J. Immunol. 166:6735–41 Mold C, Gresham HD, Du Clos TW. 2001. Serum amyloid P component and C-reactive protein mediate phagocytosis through murine Fc gamma Rs. J. Immunol. 166:1200–5 Raghavan M, Bjorkman PJ. 1996. Fc receptors and their interactions with immunoglobulins. Annu. Rev. Cell Dev. Biol. 12:181–220 Richards ML, Katz DH. 1991. Biology and chemistry of low affinity IgE receptor (Fc epsilon RII/CD23). Crit. Rev. Immunol. 11:65–86 van Spriel AB, van den Herik-Oudijk IE, van Sorge NM, Vile HA, van Strijp JA, van de Winkel JG. 1999. Effective phagocytosis and killing of Candida albicans via targeting FcgammaRI (CD64) or FcalphaRI (CD89) on neutrophils. J. Infect. Dis. 179:661–69 Ghiran I, Barbashov SF, Klickstein LB, Tas SW, Jensenius JC, Nicholson-Weller A. 2000. Complement receptor 1/CD35 is a receptor for mannan-binding lectin. J. Exp. Med. 192:1797–1808 Ehlers MR. 2000. CR3: a general purpose adhesion-recognition receptor essential for innate immunity. Microbes Infect. 2:289–94 Ross GD, Reed W, Dalzell JG, Becker SE, Hogg N. 1992. Macrophage cytoskeleton association with CR3 and CR4 regulates receptor mobility and phagocy-

30.

31.

32.

33.

34.

35.

36.

37.

38. 39.

40.

843

tosis of iC3b-opsonized erythrocytes. J. Leuk. Biol. 51:109–17 Blystone SD, Brown EJ. 1999. Integrin receptors of phagocytes. See Ref. 1, pp. 103–47 Platt N, Gordon S. 2001. Is the class A macrophage scavenger receptor (SRA) multifunctional?—The mouse’s tale. J. Clin. Invest. 108:649–54 van der Laan LJ, Dopp EA, Haworth R, Pikkarainen T, Kangas M, Elomaa O, Dijkstra CD, Gordon S, Tryggvason K, Kraal G. 1999. Regulation and functional involvement of macrophage scavenger receptor MARCO in clearance of bacteria in vivo. J. Immunol. 162:939–47 Ezekowitz RA, Sastry K, Bailly P, Warner A. 1990. Molecular characterization of the human macrophage mannose receptor: demonstration of multiple carbohydrate recognition-like domains and phagocytosis of yeasts in Cos-1 cells. J. Exp. Med. 172:1785–94 Brown GD, Gordon S. 2001. Immune recognition. A new receptor for betaglucans. Nature 413:36–37 Schiff DE, Kline L, Soldau K, Lee JD, Pugin J, Tobias PS, Ulevitch RJ. 1997. Phagocytosis of gram-negative bacteria by a unique CD14-dependent mechanism. J. Leuk. Biol. 62:786–94 Dziarski R, Tapping RI, Tobias PS. 1998. Binding of bacterial peptidoglycan to CD14. J. Biol. Chem. 273:8680–90 Nepomuceno RR, Henschen-Edman AH, Burgess WH, Tenner AJ. 1997. cDNA cloning and primary structure analysis of C1qR(P), the human C1q/MBL/SPA receptor that mediates enhanced phagocytosis in vitro. Immunity 6:119–29 Daeron M. 1997. Fc receptor biology. Annu. Rev. Immunol. 15:203–34 Ross GD. 2000. Regulation of the adhesion versus cytotoxic functions of the Mac-1/CR3/alphaMbeta2-integrin glycoprotein. Crit. Rev. Immunol. 20:197– 222 Klickstein LB, Barbashov SF, Liu T, Jack

1 Feb 2002

10:3

844

41.

Annu. Rev. Immunol. 2002.20:825-852. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

42.

43.

44.

45.

46.

47.

48.

49.

50.

AR

AR152-25.tex

UNDERHILL

¥

AR152-25.SGM

LaTeX2e(2001/05/10)

P1: IKH

OZINSKY

RM, Nicholson-Weller A. 1997. Complement receptor type 1 (CR1, CD35) is a receptor for C1q. Immunity 7:345–55 Berger M, O’Shea J, Cross AS, Folks TM, Chused TM, Brown EJ, Frank MM. 1984. Human neutrophils increase expression of C3bi as well as C3b receptors upon activation. J. Clin. Invest. 74:1566–71 Sengelov H, Kjeldsen L, Diamond MS, Springer TA, Borregaard N. 1993. Subcellular localization and dynamics of Mac-1 (alpha m beta 2) in human neutrophils. J. Clin. Invest. 92:1467–76 Jones SL, Knaus UG, Bokoch GM, Brown EJ. 1998. Two signaling mechanisms for activation of alphaM beta2 avidity in polymorphonuclear neutrophils. J. Biol. Chem. 273:10556–66 Wright SD, Griffin FM Jr. 1985. Activation of phagocytic cells’ C3 receptors for phagocytosis. J. Leuk. Biol. 38:327–39 Pommier CG, Inada S, Fries LF, Takahashi T, Frank MM, Brown EJ. 1983. Plasma fibronectin enhances phagocytosis of opsonized particles by human peripheral blood monocytes. J. Exp. Med. 157:1844–54 Caron E, Self AJ, Hall A. 2000. The GTPase Rap1 controls functional activation of macrophage integrin alphaMbeta2 by LPS and other inflammatory mediators. Curr. Biol. 10:974–78 Ehlenberger AG, Nussenzweig V. 1977. The role of membrane receptors for C3b and C3d in phagocytosis. J. Exp. Med. 145:357–71 Platt N, Haworth R, da Silva RP, Gordon S. 1999. Scavenger receptors and phagocytosis of bacteria and apoptotic cells. See Ref. 1, pp. 71–85 Hughes DA, Fraser IP, Gordon S. 1995. Murine macrophage scavenger receptor: in vivo expression and function as receptor for macrophage adhesion in lymphoid and non-lymphoid organs. Eur. J. Immunol. 25:466–73 Dunne DW, Resnick D, Greenberg J, Krieger M, Joiner KA. 1994. The type

51.

52.

53.

54.

55. 56.

57.

58.

I macrophage scavenger receptor binds to gram-positive bacteria and recognizes lipoteichoic acid. Proc. Natl. Acad. Sci. USA 91:1863–67 Peiser L, Gough PJ, Kodama T, Gordon S. 2000. Macrophage class A scavenger receptor-mediated phagocytosis of Escherichia coli: role of cell heterogeneity, microbial strain, and culture conditions in vitro. Infect. Immun. 68:1953–63 Thomas CA, Li Y, Kodama T, Suzuki H, Silverstein SC, El Khoury J. 2000. Protection from lethal gram-positive infection by macrophage scavenger receptordependent phagocytosis. J. Exp. Med. 191:147–56 Elomaa O, Kangas M, Sahlberg C, Tuukkanen J, Sormunen R, Liakka A, Thesleff I, Kraal G, Tryggvason K. 1995. Cloning of a novel bacteria-binding receptor structurally related to scavenger receptors and expressed in a subset of macrophages. Cell 80:603–9 Palecanda A, Paulauskis J, Al-Mutairi E, Imrich A, Qin G, Suzuki H, Kodama T, Tryggvason K, Koziel H, Kobzik L. 1999. Role of the scavenger receptor MARCO in alveolar macrophage binding of unopsonized environmental particles. J. Exp. Med. 189:1497–1506 Di Carlo FJ, Fiore JV. 1958. On the composition of zymosan. Science 127:756–57 Sung SS, Nelson RS, Silverstein SC. 1983. Yeast mannans inhibit binding and phagocytosis of zymosan by mouse peritoneal macrophages. J. Cell Biol. 96:160– 66 Janusz MJ, Austen KF, Czop JK. 1986. Isolation of soluble yeast beta-glucans that inhibit human monocyte phagocytosis mediated by beta-glucan receptors. J. Immunol. 137:3270–76 Giaimis J, Lombard Y, Fonteneau P, Muller CD, Levy R, Makaya Kumba M, Lazdins J, Poindron P. 1993. Both mannose and beta-glucan receptors are involved in phagocytosis of unopsonized, heat-killed Saccharomyces cerevisiae by

1 Feb 2002

10:3

AR

AR152-25.tex

AR152-25.SGM

LaTeX2e(2001/05/10)

P1: IKH

PHAGOCYTOSIS

59.

Annu. Rev. Immunol. 2002.20:825-852. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

60.

61.

62.

63.

64.

65.

murine macrophages. J. Leuk. Biol. 54: 564–71 Kruskal BA, Sastry K, Warner AB, Mathieu CE, Ezekowitz RA. 1992. Phagocytic chimeric receptors require both transmembrane and cytoplasmic domains from the mannose receptor. J. Exp. Med. 176:1673–80 Ariizumi K, Shen GL, Shikano S, Xu S, Ritter R 3rd, Kumamoto T, Edelbaum D, Morita A, Bergstresser PR, Takashima A. 2000. Identification of a novel, dendritic cell-associated molecule, dectin-1, by subtractive cDNA cloning. J. Biol. Chem. 275:20,157–67 Ariizumi K, Shen GL, Shikano S, Ritter R, 3rd, Zukas P, Edelbaum D, Morita A, Takashima A. 2000. Cloning of a second dendritic cell-associated C-type lectin (dectin-2) and its alternatively spliced isoforms. J. Biol. Chem. 275:11957–63 Willment JA, Gordon S, Brown GD. 2001. Characterisation of the human {beta}-glucan receptor and its alternatively spliced isoforms. J. Biol. Chem. 20:20 Ross GD, Cain JA, Lachmann PJ. 1985. Membrane complement receptor type three (CR3) has lectin-like properties analogous to bovine conglutinin and functions as a receptor for zymosan and rabbit erythrocytes as well as a receptor for iC3b. J. Immunol. 134:3307–15 Ross GD, Thompson RA, Walport MJ, Springer TA, Watson JV, Ward RH, Lida J, Newman SL, Harrison RA, Lachmann PJ. 1985. Characterization of patients with an increased susceptibility to bacterial infections and a genetic deficiency of leukocyte membrane complement receptor type 3 and the related membrane antigen LFA-1. Blood 66:882–90 Blystone SD, Graham IL, Lindberg FP, Brown EJ. 1994. Integrin alpha v beta 3 differentially regulates adhesive and phagocytic functions of the fibronectin receptor alpha 5 beta 1. J. Cell Biol. 127:1129–37

845

66. Blystone SD, Lindberg FP, LaFlamme SE, Brown EJ. 1995. Integrin beta 3 cytoplasmic tail is necessary and sufficient for regulation of alpha 5 beta 1 phagocytosis by alpha v beta 3 and integrin-associated protein. J. Cell Biol. 130:745–54 67. Demeure CE, Tanaka H, Mateo V, Rubio M, Delespesse G, Sarfati M. 2000. CD47 engagement inhibits cytokine production and maturation of human dendritic cells. J. Immunol. 164:2193–99 68. Oldenborg PA, Gresham HD, Lindberg FP. 2001. CD47-signal regulatory protein alpha (SIRPalpha) regulates Fcgamma and complement receptor-mediated phagocytosis. J. Exp. Med. 193:855–62 69. Gresham HD, Zheleznyak A, Mormol JS, Brown EJ. 1990. Studies on the molecular mechanisms of human neutrophil Fc receptor-mediated phagocytosis. Evidence that a distinct pathway for activation of the respiratory burst results in reactive oxygen metabolite-dependent amplification of ingestion. J. Biol. Chem. 265:7819–26 70. Senior RM, Gresham HD, Griffin GL, Brown EJ, Chung AE. 1992. Entactin stimulates neutrophil adhesion and chemotaxis through interactions between its Arg-Gly-Asp (RGD) domain and the leukocyte response integrin. J. Clin. Invest. 90:2251–57 71. Albert ML, Pearce SF, Francisco LM, Sauter B, Roy P, Silverstein RL, Bhardwaj N. 1998. Immature dendritic cells phagocytose apoptotic cells via alphavbeta5 and CD36, and cross-present antigens to cytotoxic T lymphocytes. J. Exp. Med. 188:1359–68 72. Fadok VA, Warner ML, Bratton DL, Henson PM. 1998. CD36 is required for phagocytosis of apoptotic cells by human macrophages that use either a phosphatidylserine receptor or the vitronectin receptor (alpha v beta 3). J. Immunol. 161:6250–57 73. Coppolino MG, Kong C, Mohtashami M, Schreiber AD, Brumell JH, Finlay BB,

1 Feb 2002

10:3

846

Annu. Rev. Immunol. 2002.20:825-852. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

74.

75.

76.

77.

78.

79.

80.

81.

AR

AR152-25.tex

UNDERHILL

¥

AR152-25.SGM

LaTeX2e(2001/05/10)

P1: IKH

OZINSKY

Grinstein S, Trimble WS. 2001. Requirement for N-ethylmaleimide-sensitive factor activity at different stages of bacterial invasion and phagocytosis. J. Biol. Chem. 276:4772–80 May RC, Caron E, Hall A, Machesky LM. 2000. Involvement of the Arp2/3 complex in phagocytosis mediated by FcgammaR or CR3. Nat. Cell Biol. 2:246– 48 Antonieta Cote-Velez MJ, Ortega E, Ortega A. 2001. Involvement of pp125(FAK) and p60(SRC) in the signaling through FcgammaRII-FcgammaRIII in murine macrophages. Immunol. Lett. 78:189–94 Bonilla FA, Fujita RM, Pivniouk VI, Chan AC, Geha RS. 2000. Adapter proteins SLP-76 and BLNK both are expressed by murine macrophages and are linked to signaling via Fcgamma receptors I and II/III. Proc. Natl. Acad. Sci. USA 97:1725–30 Uchida H, Kondo A, Yoshimura Y, Mazaki Y, Sabe H. 2001. PAG3/Papalpha/ KIAA0400, a GTPase-activating protein for ADP-ribosylation factor (ARF), regulates ARF6 in Fcgamma receptor-mediated phagocytosis of macrophages. J. Exp. Med. 193:955–66 Witke W, Li W, Kwiatkowski DJ, Southwick FS. 2001. Comparisons of CapG and gelsolin-null macrophages: demonstration of a unique role for CapG in receptormediated ruffling, phagocytosis, and vesicle rocketing. J. Cell Biol. 154:775–84 Breton A, Descoteaux A. 2000. Protein kinase C-alpha participates in FcgammaR-mediated phagocytosis in macrophages. Biochem. Biophys. Res. Commun. 276:472–76 Vieira OV, Botelho RJ, Rameh L, Brachmann SM, Matsuo T, Davidson HW, Schreiber A, Backer JM, Cantley LC, Grinstein S. 2001. Distinct roles of class I and class III phosphatidylinositol 3-kinases in phagosome formation and maturation. J. Cell Biol. 155:19–26 Kedzierska K, Vardaxis NJ, Jaworowski

82.

83.

84.

85.

86.

87.

88.

89.

A, Crowe SM. 2001. FcgammaR-mediated phagocytosis by human macrophages involves Hck, Syk, and Pyk2 and is augmented by GM-CSF. J. Leuk. Biol. 70:322–28 Bierne H, Gouin E, Roux P, Caroni P, Yin HL, Cossart P. 2001. A role for cofilin and LIM kinase in Listeria-induced phagocytosis. J. Cell Biol. 155:101–12 Cox D, Lee DJ, Dale BM, Calafat J, Greenberg S. 2000. A Rab11-containing rapidly recycling compartment in macrophages that promotes phagocytosis. Proc. Natl. Acad. Sci. USA 97:680–85 Hackam DJ, Botelho RJ, Sjolin C, Rotstein OD, Robinson JM, Schreiber AD, Grinstein S. 2001. Indirect role for COPI in the completion of FCgamma receptormediated phagocytosis. J. Biol. Chem. 276:18200–8 Duclos S, Diez R, Garin J, Papadopoulou B, Descoteaux A, Stenmark H, Desjardins M. 2000. Rab5 regulates the kiss and run fusion between phagosomes and endosomes and the acquisition of phagosome leishmanicidal properties in RAW 264.7 macrophages. J. Cell Sci. 113 Pt 19:3531–41 Suzuki T, Kono H, Hirose N, Okada M, Yamamoto T, Yamamoto K, Honda Z. 2000. Differential involvement of Src family kinases in Fc gamma receptormediated phagocytosis. J. Immunol. 165: 473–82 Rupper A, Grove B, Cardelli J. 2001. Rab7 regulates phagosome maturation in Dictyostelium. J. Cell Sci. 114:2449–60 Fitzer-Attas CJ, Lowry M, Crowley MT, Finn AJ, Meng F, DeFranco AL, Lowell CA. 2000. Fcgamma receptor-mediated phagocytosis in macrophages lacking the Src family tyrosine kinases Hck, Fgr, and Lyn. J. Exp. Med. 191:669–82 Gresham HD, Dale BM, Potter JW, Chang PW, Vines CM, Lowell CA, Lagenaur CF, Willman CL. 2000. Negative regulation of phagocytosis in murine macrophages by the Src kinase family

1 Feb 2002

10:3

AR

AR152-25.tex

AR152-25.SGM

LaTeX2e(2001/05/10)

P1: IKH

PHAGOCYTOSIS

Annu. Rev. Immunol. 2002.20:825-852. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

90.

91.

92.

93.

94.

95.

96.

97.

member, Fgr. J. Exp. Med. 191:515– 28 Cox D, Dale BM, Kashiwada M, Helgason CD, Greenberg S. 2001. A regulatory role for Src homology 2 domaincontaining inositol 50 -phosphatase (SHIP) in phagocytosis mediated by Fc gamma receptors and complement receptor 3 (alpha(M)beta(2); CD11b/CD18). J. Exp. Med. 193:61–71 Ohsawa K, Imai Y, Kanazawa H, Sasaki Y, Kohsaka S. 2000. Involvement of Iba1 in membrane ruffling and phagocytosis of macrophages/microglia. J. Cell Sci. 113:3073–84 Tridandapani S, Lyden TW, Smith JL, Carter JE, Coggeshall KM, Anderson CL. 2000. The adapter protein LAT enhances fcgamma receptor-mediated signal transduction in myeloid cells. J. Biol. Chem. 275:20,480–87 Bajno L, Peng XR, Schreiber AD, Moore HP, Trimble WS, Grinstein S. 2000. Focal exocytosis of VAMP3-containing vesicles at sites of phagosome formation. J. Cell Biol. 149: 697–706 Lorenzi R, Brickell PM, Katz DR, Kinnon C, Thrasher AJ. 2000. WiskottAldrich syndrome protein is necessary for efficient IgG-mediated phagocytosis. Blood 95:2943–46 Zhang J, Shehabeldin A, da Cruz LA, Butler J, Somani AK, McGavin M, Kozieradzki I, dos Santos AO, Nagy A, Grinstein S, Penninger JM, Siminovitch KA. 1999. Antigen receptor-induced activation and cytoskeletal rearrangement are impaired in Wiskott-Aldrich syndrome protein-deficient lymphocytes. J. Exp. Med. 190:1329–42 Mansfield PJ, Shayman JA, Boxer LA. 2000. Regulation of polymorphonuclear leukocyte phagocytosis by myosin light chain kinase after activation of mitogenactivated protein kinase. Blood 95:2407– 12 Chan TO, Rittenhouse SE, Tsichlis PN. 1999. AKT/PKB and other D3 phos-

98.

99.

100.

101.

102.

103.

104.

105.

847

phoinositide-regulated kinases: kinase activation by phosphoinositide-dependent phosphorylation. Annu. Rev. Biochem. 68:965–1014 Araki N, Johnson MT, Swanson JA. 1996. A role for phosphoinositide 3-kinase in the completion of macropinocytosis and phagocytosis by macrophages. J. Cell Biol. 135:1249–60 Cox D, Tseng CC, Bjekic G, Greenberg S. 1999. A requirement for phosphatidylinositol 3-kinase in pseudopod extension. J. Biol. Chem. 274:1240–47 Botelho RJ, Teruel M, Dierckman R, Anderson R, Wells A, York JD, Meyer T, Grinstein S. 2000. Localized biphasic changes in phosphatidylinositol-4,5-bisphosphate at sites of phagocytosis. J. Cell Biol. 151:1353–68 Arbibe L, Mira JP, Teusch N, Kline L, Guha M, Mackman N, Godowski PJ, Ulevitch RJ, Knaus UG. 2000. Toll-like receptor 2-mediated NF-kappa B activation requires a Rac1-dependent pathway. Nat. Immunol. 1:533–40 Chen CC, Wang JK, Lin SB. 1998. Antisense oligonucleotides targeting protein kinase C-alpha, -beta I, or -delta but not -eta inhibit lipopolysaccharide-induced nitric oxide synthase expression in RAW 264.7 macrophages: involvement of a nuclear factor kappa B-dependent mechanism. J. Immunol. 161:6206–14 Chen CC, Sun YT, Chen JJ, Chiu KT. 2000. TNF-alpha-induced cyclooxygenase-2 expression in human lung epithelial cells: involvement of the phospholipase C-gamma 2, protein kinase C-alpha, tyrosine kinase, NF-kappa B-inducing kinase, and I-kappa B kinase 1/2 pathway. J. Immunol. 165:2719–28 Bar-Sagi D, Hall A. 2000. Ras and Rho GTPases: a family reunion. Cell 103:227– 38 Chimini G, Chavrier P. 2000. Function of Rho family proteins in actin dynamics during phagocytosis and engulfment. Nat. Cell. Biol. 2:E191–6

1 Feb 2002

10:3

Annu. Rev. Immunol. 2002.20:825-852. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

848

AR

AR152-25.tex

UNDERHILL

¥

AR152-25.SGM

LaTeX2e(2001/05/10)

P1: IKH

OZINSKY

106. Caron E, Hall A. 1998. Identification of two distinct mechanisms of phagocytosis controlled by different Rho GTPases. Science 282:1717–21 107. Massol P, Montcourrier P, Guillemot JC, Chavrier P. 1998. Fc receptor-mediated phagocytosis requires CDC42 and Rac1. Embo J. 17:6219–29 108. Guillen N, Boquet P, Sansonetti P. 1998. The small GTP-binding protein RacG regulates uroid formation in the protozoan parasite Entamoeba histolytica. J. Cell Sci. 111:1729–39 109. Cox D, Chang P, Zhang Q, Reddy PG, Bokoch GM, Greenberg S. 1997. Requirements for both Rac1 and Cdc42 in membrane ruffling and phagocytosis in leukocytes. J. Exp. Med. 186:1487–94 110. Black DS, Bliska JB. 2000. The RhoGAP activity of the Yersinia pseudotuberculosis cytotoxin YopE is required for antiphagocytic function and virulence. Mol. Microbiol. 37:515–27 111. Zheleznyak A, Brown EJ. 1992. Immunoglobulin-mediated phagocytosis by human monocytes requires protein kinase C activation. Evidence for protein kinase C translocation to phagosomes. J. Biol. Chem. 267:12042–48 112. Allen LH, Aderem A. 1995. A role for MARCKS, the alpha isozyme of protein kinase C and myosin I in zymosan phagocytosis by macrophages. J. Exp. Med. 182:829–40 113. Melendez AJ, Harnett MM, Allen JM. 1999. Differentiation-dependent switch in protein kinase C isoenzyme activation by FcgammaRI, the human high-affinity receptor for immunoglobulin G. Immunology 96:457–64 114. Zheng L, Zomerdijk TP, Aarnoudse C, van Furth R, Nibbering PH. 1995. Role of protein kinase C isozymes in Fc gamma receptor-mediated intracellular killing of Staphylococcus aureus by human monocytes. J. Immunol. 155:776–84 115. Larsen EC, DiGennaro JA, Saito N, Mehta S, Loegering DJ, Mazurkiewicz

116.

117.

118.

119.

120.

121.

122.

123.

124.

JE, Lennartz MR. 2000. Differential requirement for classic and novel PKC isoforms in respiratory burst and phagocytosis in RAW 264.7 cells. J. Immunol. 165:2809–17 Shapira L, Takashiba S, Champagne C, Amar S, Van Dyke TE. 1994. Involvement of protein kinase C and protein tyrosine kinase in lipopolysaccharide-induced TNFalpha and IL-1 beta production by human monocytes. J. Immunol. 153:1818–24 Huwiler A, Pfeilschifter J. 1993. A role for protein kinase C-alpha in zymosanstimulated eicosanoid synthesis in mouse peritoneal macrophages. Eur. J. Biochem. 217:69–75 Giroux M, Descoteaux A. 2000. Cyclooxygenase-2 expression in macrophages: modulation by protein kinase C-alpha. J. Immunol. 165:3985–91 St-Denis A, Chano F, Tremblay P, StPierre Y, Descoteaux A. 1998. Protein kinase C-alpha modulates lipopolysaccharide-induced functions in a murine macrophage cell line. J. Biol. Chem. 273: 32787–92 St-Denis A, Caouras V, Gervais F, Descoteaux A. 1999. Role of protein kinase C-alpha in the control of infection by intracellular pathogens in macrophages. J. Immunol. 163:5505–11 Zhu T, Goh EL, Graichen R, Ling L, Lobie PE. 2001. Signal transduction via the growth hormone receptor. Cell Signal. 13:599–616 Kikkawa U, Kishimoto A, Nishizuka Y. 1989. The protein kinase C family: heterogeneity and its implications. Annu. Rev. Biochem. 58:31–44 Ravetch JV, Clynes RA. 1998. Divergent roles for Fc receptors and complement in vivo. Annu. Rev. Immunol. 16:421–32 Underhill DM, Ozinsky A, Hajjar AM, Stevens A, Wilson CB, Bassetti M, Aderem A. 1999. The Toll-like receptor 2 is recruited to macrophage phagosomes and discriminates between pathogens. Nature 401:811–15

1 Feb 2002

10:3

AR

AR152-25.tex

AR152-25.SGM

LaTeX2e(2001/05/10)

P1: IKH

Annu. Rev. Immunol. 2002.20:825-852. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

PHAGOCYTOSIS 125. Ozinsky A, Underhill DM, Fontenot JD, Hajjar AM, Smith KD, Wilson CB, Schroeder L, Aderem A. 2000. The repertoire for pattern recognition of pathogens by the innate immune system is defined by cooperation between toll-like receptors. Proc. Natl. Acad. Sci. USA 97:13766–71 126. Gresham HD, Graham IL, Anderson DC, Brown EJ. 1991. Leukocyte adhesiondeficient neutrophils fail to amplify phagocytic function in response to stimulation. Evidence for CD11b/CD18-dependent and -independent mechanisms of phagocytosis. J. Clin. Invest. 88:588–97 127. Gresham HD, McGarr JA, Shackelford PG, Brown EJ. 1988. Studies on the molecular mechanisms of human Fc receptor-mediated phagocytosis. Amplification of ingestion is dependent on the generation of reactive oxygen metabolites and is deficient in polymorphonuclear leukocytes from patients with chronic granulomatous disease. J. Clin. Invest. 82:1192–201 128. Graham IL, Lefkowith JB, Anderson DC, Brown EJ. 1993. Immune complexstimulated neutrophil LTB4 production is dependent on beta 2 integrins. J. Cell Biol. 120:1509–17 129. DeLeo FR, Allen LA, Apicella M, Nauseef WM. 1999. NADPH oxidase activation and assembly during phagocytosis. J. Immunol. 163:6732–40 130. Segal AW, Wientjes F, Stockley RW, Dekker LV. 1999. Components and organization of the NADPH oxidase of phagocytic cells. See Ref. 1, pp. 441–83 131. Diebold BA, Bokoch GM. 2001. Molecular basis for Rac2 regulation of phagocyte NADPH oxidase. Nat. Immunol. 2:211– 15 132. Berton G, Laudanna C, Sorio C, Rossi F. 1992. Generation of signals activating neutrophil functions by leukocyte integrins: LFA-1 and gp150/95, but not CR3, are able to stimulate the respiratory burst of human neutrophils. J. Cell Biol. 116:1007–17

849

133. Serrander L, Larsson J, Lundqvist H, Lindmark M, Fallman M, Dahlgren C, Stendahl O. 1999. Particles binding beta (2)-integrins mediate intracellular production of oxidative metabolites in human neutrophils independently of phagocytosis. Biochim. Biophys. Acta 1452:133–44 134. Grogan A, Reeves E, Keep N, Wientjes F, Totty NF, Burlingame AL, Hsuan JJ, Segal AW. 1997. Cytosolic phox proteins interact with and regulate the assembly of coronin in neutrophils. J. Cell Sci. 110:3071–81 135. Sakata A, Ida E, Tominaga M, Onoue K. 1987. Arachidonic acid acts as an intracellular activator of NADPH-oxidase in Fc gamma receptor-mediated superoxide generation in macrophages. J. Immunol. 138:4353–59 136. Wright SD, Silverstein SC. 1983. Receptors for C3b and C3bi promote phagocytosis but not the release of toxic oxygen from human phagocytes. J. Exp. Med. 158:2016–23 137. Yamamoto K, Johnston RB Jr. 1984. Dissociation of phagocytosis from stimulation of the oxidative metabolic burst in macrophages. J. Exp. Med. 159:405–16 138. Lofgren R, Serrander L, Forsberg M, Wilsson A, Wasteson A, Stendahl O. 1999. CR3, FcgammaRIIA and FcgammaRIIIB induce activation of the respiratory burst in human neutrophils: the role of intracellular Ca(2+), phospholipase D and tyrosine phosphorylation. Biochim. Biophys. Acta 1452:46–59 139. Stahl PD, Ezekowitz RA. 1998. The mannose receptor is a pattern recognition receptor involved in host defense. Curr. Opin. Immunol. 10:50–55 140. Fraser IP, Ezekowitz RA. 1999. Mannose receptor and phagocytosis. See Ref. 1, pp. 87–101. Stamford, CT: JAI 141. Astarie-Dequeker C, N’Diaye EN, Le Cabec V, Rittig MG, Prandi J, Maridonneau-Parini I. 1999. The mannose receptor mediates uptake of pathogenic and nonpathogenic mycobacteria and

1 Feb 2002

10:3

850

142.

143.

Annu. Rev. Immunol. 2002.20:825-852. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

144.

145.

146.

147.

148.

149.

150.

151.

AR

AR152-25.tex

UNDERHILL

¥

AR152-25.SGM

LaTeX2e(2001/05/10)

P1: IKH

OZINSKY

bypasses bactericidal responses in human macrophages. Infect. Immun. 67:469–77 Akira S, Takeda K, Kaisho T. 2001. Toll-like receptors: critical proteins linking innate and acquired immunity. Nat. Immunol. 2:675–80 Aderem A, Ulevitch RJ. 2000. Toll-like receptors in the induction of the innate immune response. Nature 406:782–87 Perera PY, Mayadas TN, Takeuchi O, Akira S, Zaks-Zilberman M, Goyert SM, Vogel SN. 2001. CD11b/CD18 acts in concert with CD14 and Toll-like receptor (TLR) 4 to elicit full lipopolysaccharide and taxol-inducible gene expression. J. Immunol. 166:574–81 Nigou J, Zelle-Rieser C, Gilleron M, Thurnher M, Puzo G. 2001. Mannosylated lipoarabinomannans inhibit IL-12 production by human dendritic cells: evidence for a negative signal delivered through the mannose receptor. J. Immunol. 166:7477–85 Celli J, Olivier M, Finlay BB. 2001. Enteropathogenic Escherichia coli mediates antiphagocytosis through the inhibition of PI 3-kinase-dependent pathways. Embo J. 20:1245–58 Rosqvist R, Forsberg A, Rimpilainen M, Bergman T, Wolf-Watz H. 1990. The cytotoxic protein YopE of Yersinia obstructs the primary host defence. Mol. Microbiol. 4:657–67 Ohl ME, Miller SI. 2001. Salmonella: a model for bacterial pathogenesis. Annu. Rev. Med. 52:259–74 Fallman M, Andersson K, Hakansson S, Magnusson KE, Stendahl O, Wolf-Watz H. 1995. Yersinia pseudotuberculosis inhibits Fc receptor-mediated phagocytosis in J774 cells. Infect. Immun. 63:3117–24 Ruckdeschel K, Roggenkamp A, Schubert S, Heesemann J. 1996. Differential contribution of Yersinia enterocolitica virulence factors to evasion of microbicidal action of neutrophils. Infect. Immun. 64:724–33 Black DS, Bliska JB. 1997. Identifica-

152.

153.

154.

155.

156.

157.

158.

159.

tion of p130Cas as a substrate of Yersinia YopH (Yop51), a bacterial protein tyrosine phosphatase that translocates into mammalian cells and targets focal adhesions. EMBO J. 16:2730–44 Andersson K, Carballeira N, Magnusson KE, Persson C, Stendahl O, WolfWatz H, Fallman M. 1996. YopH of Yersinia pseudotuberculosis interrupts early phosphotyrosine signalling associated with phagocytosis. Mol. Microbiol. 20:1057–69 Persson C, Carballeira N, Wolf-Watz H, Fallman M. 1997. The PTPase YopH inhibits uptake of Yersinia, tyrosine phosphorylation of p130Cas and FAK, and the associated accumulation of these proteins in peripheral focal adhesions. EMBO J. 16:2307–18 Valdivia RH, Falkow S. 1997. Fluorescence-based isolation of bacterial genes expressed within host cells. Science 277: 2007–11 Shea JE, Hensel M, Gleeson C, Holden DW. 1996. Identification of a virulence locus encoding a second type III secretion system in Salmonella typhimurium. Proc. Natl. Acad. Sci. USA 93:2593–97 Ochman H, Soncini FC, Solomon F, Groisman EA. 1996. Identification of a pathogenicity island required for Salmonella survival in host cells. Proc. Natl. Acad. Sci. USA 93:7800–4 Vazquez-Torres A, Xu Y, Jones-Carson J, Holden DW, Lucia SM, Dinauer MC, Mastroeni P, Fang FC. 2000. Salmonella pathogenicity island 2-dependent evasion of the phagocyte NADPH oxidase. Science 287:1655–58 Uchiya K, Barbieri MA, Funato K, Shah AH, Stahl PD, Groisman EA. 1999. A Salmonella virulence protein that inhibits cellular trafficking. EMBO J. 18:3924–33 Miao EA, Miller SI. 2000. A conserved amino acid sequence directing intracellular type III secretion by Salmonella typhimurium. Proc. Natl. Acad. Sci. USA 97:7539–44

1 Feb 2002

10:3

AR

AR152-25.tex

AR152-25.SGM

LaTeX2e(2001/05/10)

P1: IKH

Annu. Rev. Immunol. 2002.20:825-852. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

PHAGOCYTOSIS 160. Russell DG. 2001. Mycobacterium tuberculosis: here today, and here tomorrow. Nat. Rev. Mol. Cell Biol. 2:569–77 161. Ullrich HJ, Beatty WL, Russell DG. 1999. Direct delivery of procathepsin D to phagosomes: implications for phagosome biogenesis and parasitism by Mycobacterium. Eur. J. Cell Biol. 78:739–48 162. Clemens DL, Lee BY, Horwitz MA. 2000. Deviant expression of Rab5 on phagosomes containing the intracellular pathogens Mycobacterium tuberculosis and Legionella pneumophila is associated with altered phagosomal fate. Infect. Immun. 68:2671–84 163. Via LE, Deretic D, Ulmer RJ, Hibler NS, Huber LA, Deretic V. 1997. Arrest of mycobacterial phagosome maturation is caused by a block in vesicle fusion between stages controlled by rab5 and rab7. J. Biol. Chem. 272:13,326–31 164. Ferrari G, Langen H, Naito M, Pieters J. 1999. A coat protein on phagosomes involved in the intracellular survival of mycobacteria. Cell 97:435–47 165. Malik ZA, Iyer SS, Kusner DJ. 2001. Mycobacterium tuberculosis phagosomes exhibit altered calmodulin-dependent signal transduction: contribution to inhibition of phagosome-lysosome fusion and intracellular survival in human macrophages. J. Immunol. 166:3392–3401 166. Malik ZA, Denning GM, Kusner DJ. 2000. Inhibition of Ca(2+) signaling by Mycobacterium tuberculosis is associated with reduced phagosome-lysosome fusion and increased survival within human macrophages. J. Exp. Med. 191:287–302 167. Reiner SL, Zheng S, Wang ZE, Stowring L, Locksley RM. 1994. Leishmania promastigotes evade interleukin 12 (IL-12) induction by macrophages and stimulate a broad range of cytokines from CD4+ T cells during initiation of infection. J. Exp. Med. 179:447–56 168. McDowell MA, Sacks DL. 1999. Inhibition of host cell signal transduction by Leishmania: observations relevant to the

169.

170.

171.

172.

173.

174.

175.

176.

851

selective impairment of IL-12 responses. Curr. Opin. Microbiol. 2:438–43 Means TK, Wang S, Lien E, Yoshimura A, Golenbock DT, Fenton MJ. 1999. Human toll-like receptors mediate cellular activation by Mycobacterium tuberculosis. J. Immunol. 163:3920–27 Underhill DM, Ozinsky A, Smith KD, Aderem A. 1999. Toll-like receptor-2 mediates mycobacteria-induced proinflammatory signaling in macrophages. Proc. Natl. Acad. Sci. USA 96:14459–63 Orth K, Xu Z, Mudgett MB, Bao ZQ, Palmer LE, Bliska JB, Mangel WF, Staskawicz B, Dixon JE. 2000. Disruption of signaling by Yersinia effector YopJ, a ubiquitin-like protein protease. Science 290:1594–97 Orth K, Palmer LE, Bao ZQ, Stewart S, Rudolph AE, Bliska JB, Dixon JE. 1999. Inhibition of the mitogen-activated protein kinase kinase superfamily by a Yersinia effector. Science 285:1920– 23 Byrd-Leifer CA, Block EF, Takeda K, Akira S, Ding A. 2001. The role of MyD88 and TLR4 in the LPS-mimetic activity of Taxol. Eur. J. Immunol. 31:2448– 57 Campos MA, Almeida IC, Takeuchi O, Akira S, Valente EP, Procopio DO, Travassos LR, Smith JA, Golenbock DT, Gazzinelli RT. 2001. Activation of Tolllike receptor-2 by glycosylphosphatidylinositol anchors from a protozoan parasite. J. Immunol. 167:416–23 Brightbill HD, Libraty DH, Krutzik SR, Yang RB, Belisle JT, Bleharski JR, Maitland M, Norgard MV, Plevy SE, Smale ST, Brennan PJ, Bloom BR, Godowski PJ, Modlin RL. 1999. Host defense mechanisms triggered by microbial lipoproteins through toll-like receptors. Science 285:732–36 Thoma-Uszynski S, Stenger S, Takeuchi O, Ochoa MT, Engele M, Sieling PA, Barnes PF, Rollinghoff M, Bolcskei PL, Wagner M, Akira S, Norgard MV, Belisle

1 Feb 2002

10:3

852

AR

AR152-25.tex

UNDERHILL

¥

AR152-25.SGM

P1: IKH

OZINSKY

JT, Godowski PJ, Bloom BR, Modlin RL. 2001. Induction of direct antimicrobial activity through mammalian toll-like receptors. Science 291:1544–47 177. Means TK, Jones BW, Schromm AB, Shurtleff BA, Smith JA, Keane J, Golenbock DT, Vogel SN, Fenton MJ. 2001. Differential effects of a Toll-like recep-

Annu. Rev. Immunol. 2002.20:825-852. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

LaTeX2e(2001/05/10)

tor antagonist on Mycobacterium tuberculosis-induced macrophage responses. J. Immunol. 166:4074–82 178. Pabst MJ, Johnston RB Jr. 1980. Increased production of superoxide anion by macrophages exposed in vitro to muramyl dipeptide or lipopolysaccharide. J. Exp. Med. 151:101–14

P1: FRK

February 1, 2002

12:47

Annual Reviews

AR152-FM

Annual Review of Immunology Volume 20, 2002

Annu. Rev. Immunol. 2002.20:825-852. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

CONTENTS FRONTISPIECE, Charles A. Janeway, Jr. A TRIP THROUGH MY LIFE WITH AN IMMUNOLOGICAL THEME, Charles A. Janeway, Jr.

xiv 1

THE B7 FAMILY OF LIGANDS AND ITS RECEPTORS: NEW PATHWAYS FOR COSTIMULATION AND INHIBITION OF IMMUNE RESPONSES, Beatriz M. Carreno and Mary Collins

29

MAP KINASES IN THE IMMUNE RESPONSE, Chen Dong, Roger J. Davis, and Richard A. Flavell

PROSPECTS FOR VACCINE PROTECTION AGAINST HIV-1 INFECTION AND AIDS, Norman L. Letvin, Dan H. Barouch, and David C. Montefiori T CELL RESPONSE IN EXPERIMENTAL AUTOIMMUNE ENCEPHALOMYELITIS (EAE): ROLE OF SELF AND CROSS-REACTIVE ANTIGENS IN SHAPING, TUNING, AND REGULATING THE AUTOPATHOGENIC T CELL REPERTOIRE, Vijay K. Kuchroo, Ana C. Anderson, Hanspeter Waldner, Markus Munder, Estelle Bettelli, and Lindsay B. Nicholson

55 73

101

NEUROENDOCRINE REGULATION OF IMMUNITY, Jeanette I. Webster, Leonardo Tonelli, and Esther M. Sternberg

125

MOLECULAR MECHANISM OF CLASS SWITCH RECOMBINATION: LINKAGE WITH SOMATIC HYPERMUTATION, Tasuku Honjo, Kazuo Kinoshita, and Masamichi Muramatsu

165

INNATE IMMUNE RECOGNITION, Charles A. Janeway, Jr. and Ruslan Medzhitov

KIR: DIVERSE, RAPIDLY EVOLVING RECEPTORS OF INNATE AND ADAPTIVE IMMUNITY, Carlos Vilches and Peter Parham ORIGINS AND FUNCTIONS OF B-1 CELLS WITH NOTES ON THE ROLE OF CD5, Robert Berland and Henry H. Wortis E PROTEIN FUNCTION IN LYMPHOCYTE DEVELOPMENT, Melanie W. Quong, William J. Romanow, and Cornelis Murre

197 217 253 301

LYMPHOCYTE-MEDIATED CYTOTOXICITY, John H. Russell and Timothy J. Ley

SIGNAL TRANSDUCTION MEDIATED BY THE T CELL ANTIGEN x

323

P1: FRK

Annu. Rev. Immunol. 2002.20:825-852. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

February 1, 2002

12:47

Annual Reviews

AR152-FM

CONTENTS

xi

RECEPTOR: THE ROLE OF ADAPTER PROTEINS, Lawrence E. Samelson INTERACTION OF HEAT SHOCK PROTEINS WITH PEPTIDES AND ANTIGEN PRESENTING CELLS: CHAPERONING OF THE INNATE AND ADAPTIVE IMMUNE RESPONSES, Pramod Srivastava CHROMATIN STRUCTURE AND GENE REGULATION IN THE IMMUNE SYSTEM, Stephen T. Smale and Amanda G. Fisher PRODUCING NATURE’S GENE-CHIPS: THE GENERATION OF PEPTIDES FOR DISPLAY BY MHC CLASS I MOLECULES, Nilabh Shastri, Susan

371

Schwab, and Thomas Serwold

395 427

463

THE IMMUNOLOGY OF MUCOSAL MODELS OF INFLAMMATION, Warren Strober, Ivan J. Fuss, and Richard S. Blumberg T CELL MEMORY, Jonathan Sprent and Charles D. Surh

GENETIC DISSECTION OF IMMUNITY TO MYCOBACTERIA: THE HUMAN MODEL, Jean-Laurent Casanova and Laurent Abel ANTIGEN PRESENTATION AND T CELL STIMULATION BY DENDRITIC CELLS, Pierre Guermonprez, Jenny Valladeau, Laurence Zitvogel, Clotilde Th´ery, and Sebastian Amigorena

495 551 581

621

NEGATIVE REGULATION OF IMMUNORECEPTOR SIGNALING, Andr´e Veillette, Sylvain Latour, and Dominique Davidson

669

CPG MOTIFS IN BACTERIAL DNA AND THEIR IMMUNE EFFECTS, Arthur M. Krieg

PROTEIN KINASE Cθ

709 IN

T CELL ACTIVATION, Noah Isakov and Amnon

Altman

RANK-L AND RANK: T CELLS, BONE LOSS, AND MAMMALIAN EVOLUTION, Lars E. Theill, William J. Boyle, and Josef M. Penninger PHAGOCYTOSIS OF MICROBES: COMPLEXITY IN ACTION, David M. Underhill and Adrian Ozinsky

761 795 825

STRUCTURE AND FUNCTION OF NATURAL KILLER CELL RECEPTORS: MULTIPLE MOLECULAR SOLUTIONS TO SELF, NONSELF DISCRIMINATION, Kannan Natarajan, Nazzareno Dimasi, Jian Wang, Roy A. Mariuzza, and David H. Margulies

853

INDEXES Subject Index Cumulative Index of Contributing Authors, Volumes 1–20 Cumulative Index of Chapter Titles, Volumes 1–20

ERRATA An online log of corrections to Annual Review of Immunology chapters (if any, 1997 to the present) may be found at http://immunol.annualreviews.org/

887 915 925

8 Feb 2002

10:14

AR

AR152-26.tex

AR152-26.SGM

LaTeX2e(2001/05/10) P1: IKH 10.1146/annurev.immunol.20.100301.064812

Annu. Rev. Immunol. 2002. 20:853–85 DOI: 10.1146/annurev.immunol.20.100301.064812

STRUCTURE AND FUNCTION OF NATURAL KILLER CELL RECEPTORS: Multiple Molecular Solutions to

Annu. Rev. Immunol. 2002.20:853-885. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

Self, Nonself Discrimination∗

Kannan Natarajan1, Nazzareno Dimasi2, Jian Wang1, Roy A. Mariuzza2,3, and David H. Margulies1,3 1

Molecular Biology Section, Laboratory of Immunology, NIAID, NIH, Bethesda, Maryland 20892-1892; e-mail: [email protected], [email protected], [email protected] 2 Center for Advanced Research in Biotechnology, Keck Laboratory of Structural Biology, University of Maryland Biotechnology Institute, 9600 Gudelsky Drive, Rockville, Maryland 20850; e-mail: [email protected], [email protected]

Key Words natural killer (NK) cells, NK receptors, missing self, inhibitory and activating receptors, T cell receptors (TCR), MHC, MICA, RAE-1, ULBP ■ Abstract In contrast to T cell receptors, signal transducing cell surface membrane molecules involved in the regulation of responses by cells of the innate immune system employ structures that are encoded in the genome rather than generated by somatic recombination and that recognize either classical MHC-I molecules or their structural relatives (such as MICA, RAE-1, or H-60). Considerable progress has recently been made in our understanding of molecular recognition by such molecules based on the determination of their three-dimensional structure, either in isolation or in complex with their MHC-I ligands. Those best studied are the receptors that are expressed on natural killer (NK) cells, but others are found on populations of T cells and other hematopoietic cells. These molecules fall into two major structural classes, those of the immunoglobulin superfamily (KIRs and LIRs) and of the C-type lectin-like family (Ly49, NKG2D, and CD94/NKG2). Here we summarize, in a functional context, the structures of the murine and human molecules that have recently been determined, with emphasis on how they bind different regions of their MHC-I ligands, and how this allows the discrimination of tumor or virus-infected cells from normal cells of the host.

INTRODUCTION The adaptive immune system of vertebrates enables the development of an exquisitely specific and powerful immune response that targets the antigenic and pathogenic assault while minimizing collateral damage to host tissues. Antigen-specific ∗ The US Government has the right to retain a nonexclusive, royalty-free license in and to any copyright covering this paper.

853

4 Feb 2002

12:4

Annu. Rev. Immunol. 2002.20:853-885. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

854

AR

AR152-26.tex

AR152-26.SGM

LaTeX2e(2001/05/10)

P1: IKH

NATARAJAN ET AL.

receptors involved in adaptive immunity, notably antibodies, B cell receptors (BCRs) and T cell receptors (TCRs), are highly discriminatory, are encoded by somatically rearranged genes, are clonally distributed on B and T lymphocytes, and recognize antigens, pathogens, and microbes or degradative fragments of their proteins. Several evolutionarily ancient families of recognition molecules co-exist in higher vetebrates with this adaptive recognition system. These include members of the immunoglobulin (Ig) superfamily distinct from antibodies and TCRs (1), the C-type lectin family of receptors (2, 3), and the Toll-like receptors (4). A general feature of these molecular families appears to be their ability to detect common molecular structures found on bacteria or other pathogens and to signal cells of the host to initiate various complex cellular inflammatory pathways. More recent products of the evolution of the C-type lectins and Ig families are the C-type lectin-like and Ig-like receptors found on NK cells that represent molecules that do not directly recognize pathogens per se but instead monitor the quantitative expression of cell surface molecules that are dysregulated as a consequence of pathologic changes within the host cell. These receptors do not undergo somatic rearrangement; they show limited diversity and are not highly specific for their ligands. The ligands recognized by these latter receptors are major histocompatibility complex class I (MHC-I) molecules and their homologs. Although our understanding of the structure and function of MHC-I specific receptors on NK cells lags behind a comparable understanding of antigen-specific receptors on T and B lymphocytes, a flurry of recent publications begins to shed some light on the functional, biochemical, and structural basis of the ligand preferences of these receptors. An important theme emerging from these recent studies is that particular NK receptors solve the problem of molecular recognition of MHC-I-related molecules in structurally distinct ways. Different families of NK receptors bind distinct molecular surface sites of their MHC-I or MHC-Ib ligands but preserve the functional ability to monitor dynamic changes in ligand expression secondary to viral infection or oncogenic transformation. The ability to sense such changes permits the NK cell to respond vigorously to such insults and to initiate not only NK cytotoxicity, but also a general inflammatory response. NK cells have been defined traditionally as a population of large, granular lymphocytes that displays spontaneous cytolysis in vitro toward a variety of cell lines. Cytolysis is inhibited when appropriate MHC-I molecules are expressed by the target cell, and it is enhanced when targets are deficient in MHC-I expression, an observation that led to the “missing self” hypothesis of NK function (5–7). The hypothesis proposes that interactions between an NK cell and a target cell will lead to lysis of the target unless the target cell expresses MHC-I molecules that can engage specific inhibitory receptors on the NK cell. Engagement of such inhibitory receptors turns off the signaling cascades that trigger cytolysis. Conversely, cells that have lost normal MHC-I expression (missing self ), as seen in tumors or in microbially infected cells, are consequently unable to deliver inhibitory signals to the NK cell and thus become susceptible to lysis. Formal demonstration of the role of MHC-I in protection of target cells from NK-mediated lysis was achieved by transfecting MHC-I alleles into otherwise MHC-I deficient, and therefore

4 Feb 2002

12:4

AR

AR152-26.tex

AR152-26.SGM

LaTeX2e(2001/05/10)

P1: IKH

Annu. Rev. Immunol. 2002.20:853-885. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

NK CELL RECEPTORS

855

NK-susceptible, targets. Thus, transfected expression of human MHC-I (HLA-A, -B, or -C) alleles resulted in protection of targets from lysis by defined NK cell subpopulations (8–10). Similarly in the mouse, transfected (11) or transgenic (12) expression of MHC-I (H-2) protected target cells from NK-mediated lysis. On the NK side of the interaction, cDNAs encoding MHC-I sensitive inhibitory receptors were initially identified by expression cloning and shown to belong to either the Ig-like family of proteins in the human (13) or the C-type lectin-like family in the mouse (14). Beginning with these studies, a bewildering array of MHC-I specific receptors has been subsequently identified on both human and mouse NK cells (15, 16). An observation not entirely explained by the missing self hypothesis is that some receptors are activating rather than inhibitory (16). In addition to their role in in vitro cytolysis, MHC-I specific inhibitory and activating receptors on NK cells can be invoked to explain the phenomenon of hybrid resistance in mice, in which, contrary to the laws of transplantation, F1 offspring of MHC-disparate parents acutely reject bone marrow grafts from either parent (17, 18). Rejection is mediated by NK cells (19) and can be modulated by antibodies to inhibitory receptors on NK cells (20). In addition to receptors specific for classical or nonclassical MHC-I proteins, a number of other surface molecules have been implicated in the regulation of NK cell cytolytic activity and cytokine production (21). These molecules include CD16 (22), CD69 (23), CD44 (24), NKR-P1 (25), CD2 (26), 2B4 (27, 28), DNAM-1 (29), NKp30 (30), NKp44 (31, 32) and NKp46 (33), which appear to contribute to NK cell activation, and the inhibitory receptor p75/AIRM1 (34). However, for only a few of these receptors are the physiological ligands known: IgG Fc for CD16 (35), human CD58 (36) or rat CD48 (37) for CD2, and CD48 for 2B4 (38). Of particular interest are NKp30, NKp44, and NKp46, which are thought to recognize certain tumors and to be the main activating receptors for human NK cells. Recently, NKp46 has also been shown to recognize cells infected by influenza virus, possibly through a direct interaction with the viral hemagglutinin (38a). The balance between inhibitory signals from receptors specific for MHC-I and stimulatory signals mediated by a variety of activating receptors ultimately determines the outcome of an NK cell–target cell encounter. In addition to the function of these receptors on NK cells, mounting evidence suggests an important regulatory role when NK receptors are expressed on T cells (39–42). Also, receptors related in amino acid sequence to the C-type lectin-like family, such as DEC-205 (43), DCSIGN (44), and DECTIN-1 (45, 46), expressed on specialized antigen presenting cells such as dendritic cells, almost certainly play a role in regulating early steps in the immune response. In the past several years, a number of laboratories have successfully determined the X-ray crystallographic structure of representative receptors either in isolation or complexed with their biologically relevant ligands. These include the human killer cell Ig-like receptors (KIRs), KIR2DL1 (47), KIR2DL2 (48), KIR2DL3 (49), KIR2DL2 in complex with HLA-Cw3 (50), KIR2DL1 in complex with HLACw4 (51), and the leukocyte Ig-like receptor (LIR-1/ILT-2) (52). The structure of

4 Feb 2002

12:4

856

AR

AR152-26.tex

AR152-26.SGM

LaTeX2e(2001/05/10)

P1: IKH

NATARAJAN ET AL.

Annu. Rev. Immunol. 2002.20:853-885. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

receptors of the C-type lectin-like family, human CD94 (53), human CD69 (54, 55) and mouse NKG2D (56) have been reported. Examples of C-type lectin-like receptor/MHC complexes include Ly49A complexed with the MHC-I molecule H-2Dd (57), human NKG2D in complex with its MHC-I like ligands MICA (58) and ULBP (S Radaev, B Rostro, AG Brooks, M Colonna, PD Sun, submitted), and murine NKG2D bound to the MHC-I homolog RAE-1 (P Li, G McDermott, RK Strong, submitted). These structures reveal both common and unique aspects of the mode of ligand recognition and demonstrate the multiplicity of molecular solutions that have evolved to accomplish the task of ligand recognition.

HUMAN KIRs On human NK cells, members of the KIR family (also referred to as CD158, see below), mediate recognition of various HLA-A, -B, and -C alleles. These receptors are type I (extracellular amino terminus) membrane proteins that contain either two or three extracellular Ig-like domains (13, 59), and hence are designated KIR2D or KIR3D receptors, respectively. (The nomenclature of the human activating and inhibitory receptors for MHC molecules is in flux. We have summarized the designations currently in wide use, as well as the recent CD classification in Table 1, and we use here the descriptive KIR, noting the CD name when appropriate.) The cytoplasmic domains of the KIRs can be either long (L) or short (S), corresponding to their function as inhibitory or activating receptors, respectively. Inhibitory KIRs contain one or two ITIM sequences (V/IxYxxL/V) in their cytoplasmic domains that, when tyrosine phosphorylated, recruit and activate SHP-1 phosphatase (60– 62), leading to inhibition of signaling. Activating receptors, on the other hand, do not directly signal but instead must associate noncovalently (via a salt bridge linking the transmembrane domains) with other molecules containing immunoreceptor tyrosine–based activation motifs (ITAMs) that serve as signal transduction units. Different activating receptors seem to exploit different adaptors (21). The short human KIRs exploit a membrane-anchored, homodimeric adaptor molecule (DAP12 or KARAP) that contains cytoplasmic ITAM sequences that activate via the syk and ZAP-70 pathways (63, 64). Other activating receptors take advantage of FcεRIγ homodimers (mouse NKR-P1 and the non-KIR activating receptor, human NKp46), FcεRIγ /CD3ζ heterodimers (human NKp46), or CD3ζ homodimers (human NKp46) (16). Also, the homodimeric C-type lectin-like activation receptor NKG2D is associated with DAP10 (also known as KAP10) and contains a YINM motif, which may act through phosphoinositide-3 kinase (PI3K) rather than an ITAM (65, 66). Inhibitory and activating receptors are often co-expressed on individual NK cells, and in such situations the inhibitory signal dominates the activating one (62). Expression of inhibitory KIRs also occurs on T cell subpopulations (67, 68) where they have been shown to modulate T cell functions (69). Similar signaling modes for activating and inhibitory receptors hold for murine molecules as well (see below).

4 Feb 2002

12:4

AR

AR152-26.tex

AR152-26.SGM

LaTeX2e(2001/05/10)

P1: IKH

NK CELL RECEPTORS

857

TABLE 1 CD nomenclature for activating and inhibitory human NK receptors

Annu. Rev. Immunol. 2002.20:853-885. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

Common names KIR3DL22 KIR2DS2 KIR2DS4 KIR2DS1 KIR2DS5 KIR2DL5 KIR3DL1/S1 KIR2DL4 KIR2DS6 KIR2DL1 KIR2DL2/L3 KIR3DL7 ILT10 ILT9 ILT3 ILT2 ILT1 ILT7 ILT11 ILT6 ILT4 ILT8 ILT5

HGNC1 Nomenclature

CD designation

p140, NKAT4/4a/4b p50.2 p50.3, NKAT8 p50.1 NKAT9

CD158k CD158j CD158i CD158h CD158g CD158f CD158e1/e2 CD158d CD158c CD158a CD158b1/b2 CD158z CD85m CD85l CD85k CD85j CD85i CD85h CD85g CD85f CD85e CD85d CD85c CD85b CD85a

p70, NKAT3/NKAT10 KIRX NKAT1, p58.1 p58.2/p58.3, NKAT6/NKAT2 KIRC1

LIR5 LIR1, MIR7 LIR6 LIR7

LILRB4 LILRB1 LILRA1 LILRA2

LIR4 LIR2, MIR10 LIR8

LILRA3 LILRB2 LILRB5

LIR3

LILRB3

(178) http://www.gene.ucl.ac.uk/nomenclature/genefamily/lilr.html http://www.gene.ucl.ac.uk/nomenclature/genefamily/kir.html http://www.ncbi.nlm.nih.gov/prow/guide/679664748 g.htm 1

HGNC, HUGO (the human genome organization) gene nomenclature committee.

2

KIR and ILT genes are located on human chromosome 19qter and 19q13.42 respectively.

Ligand Specificities of KIR Molecules The ligand specificities of inhibitory KIRs were initially identified by correlating expression of particular receptors with the inability to lyse target cells of defined HLA haplotype. Alleles of all three MHC-I molecules, HLA-A, -B, and -C can confer protection from lysis by NK clones. In an exhaustive analysis of the cytotoxicity of a panel of two hundred NK clones derived from four separate donors, it was shown that HLA-A, -B, and -C molecules can each confer protection; both self and allogeneic MHC-I are recognized; and individual NK clones recognize multiple MHC-I alleles (70). In general, KIR3D receptors recognize HLA-A and

4 Feb 2002

12:4

Annu. Rev. Immunol. 2002.20:853-885. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

858

AR

AR152-26.tex

AR152-26.SGM

LaTeX2e(2001/05/10)

P1: IKH

NATARAJAN ET AL.

-B alleles, whereas KIR2D receptors recognize HLA-C alleles. KIR2D receptors can be subdivided into two groups based on their ability to discriminate between those HLA-C alleles with either Lys80 (e.g., HLA-Cw2, -Cw4, -Cw5, -Cw6) or Asn80 (e.g., HLA-Cw1, -Cw3, -Cw7, -Cw8) in the MHC α1helix (10). In order to investigate in greater detail the interaction between KIRs and their MHC ligands, several groups have engineered recombinant, soluble versions of KIRs for direct binding studies. These have primarily focused on the interaction of KIR2D receptors with HLA-C. Fusion proteins consisting of the two extracellular Ig-like domains of KIR2DL2 linked to the amino terminus of an Ig constant region were expressed and purified. The recombinant molecules specifically stained only those transfectants expressing the cognate HLA-C allele and did not stain either the MHC-I-deficient parental cell line or an HLA-B transfectant (71, 72). Other groups have successfully used bacterial expression systems to obtain soluble KIRs and HLA-C for binding analysis (73–75), thus demonstrating that carbohydrates on either the receptor or the ligand are not necessary for specific binding. Surface plasmon resonance studies of the interaction between KIR2Ds of the two different HLA-C specificities and their HLA-C ligands indicate that no other cell surface molecules are required. The binding progress curves reveal extremely rapid association and dissociation kinetics resembling the interactions between adhesion molecules and their ligands, with calculated half-lives of less than 0.5 seconds for KIR2D/HLA-C complexes (76). Recognition by KIR3D receptors of HLA-B molecules also is dependent on the allelic residue at position 80 of the MHC-1 α1helix. A subset of KIR3Dexpressing NK clones is inhibited by targets expressing HLA-B alleles that have Ile80 (77, 78), and another subset is inhibited by HLA-B alleles in which Ile80 is replaced by threonine (79). Binding studies indicate that all three domains of KIR3D molecules are needed for interacton with HLA-B (80). It thus appears that both HLA-B and -C specific KIRs share a common mode of binding to their respective ligands involving the region around position 80 of the MHC-I α1helix. While the KIRs described above are monomers, a KIR (known as KIR3DL2, CD158k, p140 or NKAT4) that appears to be expressed as a disulfide-linked homodimer, with a molecular weight of approximately 140 kDa under nonreducing conditions, has been described that specifically recognizes HLA-A3 (81, 82). The state of knowledge of the ligand specificity of the activating KIRs is somewhat limited. However, some experiments indicate that KIR2DS1 (CD158h, p50.1) interacts weakly with HLA-Cw4 (83). Although a major function of KIRs is to scan target cells for normal expression of MHC-I molecules, recognition is not entirely independent of the bound peptide. Substitutions at positions 7 and 8 of nonamer peptides disrupt KIR recognition of HLA-B and HLA-C (84), in both binding and functional assays. Since these residues at the carboxyl end of the bound peptide lie in the vicinity of MHC-I position 80, such peptide preference is a consequence of the location of the KIR footprint on the MHC/peptide complex, as described below. In support of the view that peptide is unlikely to play a major role, “empty” HLA-C molecules on the

8 Feb 2002

10:58

AR

AR152-26.tex

AR152-26.SGM

LaTeX2e(2001/05/10)

P1: IKH

NK CELL RECEPTORS

859

cell surface of HLA-C-transfected RMA-S (TAP-defective) cells grown at 25◦ C, effectively provide protection from NK-mediated lysis (85). However, it is unclear whether such molecules are truly empty or are transiently stabilized by exogenous, TAP-independent peptides.

Annu. Rev. Immunol. 2002.20:853-885. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

The Structure of KIRs Significant advances in our understanding of the molecular basis of the specificity of KIR2D receptors for their HLA-C ligands have been provided by the crystal structures of KIR2D molecules, both free (47–49) and in complex with their respective HLA-C ligands (50, 51). KIR2D molecules have two tandem amino terminal domains, designated D1 and D2, that are connected by a linker of three to five amino acids (Figure 1A). The structural relationship to Ig superfamily proteins (86) is evident from the disposition of the two anti-parallel β-sheets forming each domain in which a β-sheet of three (in D2) or four (in D1) anti-parallel strands (ABE or ABED, repectively) packs against another β-sheet of four antiparallel strands (C0 CFG) with one extra A0 strand. The A0 strand pairs with the G strand by switching at a cis-proline from one β-sheet to the other (Figure 1A). Amino acid sequence alignment of the three KIR2D molecules that have been studied crystallographically reveals how closely related these are, and how relatively few polymorphic residues can significantly affect allelic specificity (Figure 2A). The topology of the domains and their relative orientation resemble those found in hematopoietic receptors such as human growth hormone receptor (87), erythropoietin receptor (88), and prolactin receptor (89). The angle between the D1 and D2 domains is different in each of the three KIRs, approximately 60◦ in KIR2DL1 (47), 70◦ in KIR2DL3 (49), and 80◦ in KIR2DL2 (48). This indicates the potential for significant variability of the interdomain angle, despite a relatively conserved set of hydrophobic residues in the hinge region and a large ˚ 2 between D1 and D2. The acuteness buried surface area of approximately 1000 A of the interdomain angles distinguishes KIRs from two other multidomain Ig-type receptors, CD2 (90) and CD4 (91, 92), in which the domain orientations are almost linear.

Structure of KIR/MHC-I Complexes Site-directed mutagenesis of KIR2D molecules has identified residue 44, which is methionine in KIR2DL1 and lysine in KIR2DL2, as critical for receptor specificity (93). Residues 45 and 70 have also been implicated (83, 93, 94). These residues are located in a series of loops found on the outer surface of the interdomain elbow region, which therefore constitutes the putative ligand binding site of KIR molecules. The KIR binding site on the MHC-I was localized by site-directed mutagenesis to the region around position 80 of the HLA-C α1helix and the following loop (74). These predictions were amply confirmed by the crystal structure of the KIR2DL2/HLA-Cw3 (50) and KIR2DL1/HLA-Cw4 complexes (51). KIR2DL1 and KIR2DL2 bind their respective HLA-C ligands through the α1 and α2 helices

8 Feb 2002

11:3

Annu. Rev. Immunol. 2002.20:853-885. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

860

AR

AR152-26.tex

AR152-26.SGM

LaTeX2e(2001/05/10)

P1: IKH

NATARAJAN ET AL.

and the C-terminal portion of the bound peptide (Figure 3). Thus, the axis of the D1D2 structure lies approximately orthogonal to the axis of the bound peptide and the MHC α1helix, aligning the D1 domain of the KIR with the α1helix, and D2 with α2. This orientation resembles the docking mode of TCRs onto MHC-I (95–97) but is completely distinct from that of the Ly49A NK receptor (see below and Figure 6). In both the KIR2DL1/HLA-Cw4 and KIR2DL2/HLA-Cw3 complexes, the KIR contacts HLA-C through six loops near its interdomain hinge region (Figures 1 and 2). Loops A0 B, CC0 and EF of the D1 domain contact the α1 helix of HLA-C. The hinge and loops BC and FG of D2 contact α2. Significantly, the corresponding six loops of human growth hormone receptor (87), erythropoietin receptor (98), and prolactin receptor (89) also mediate ligand recognition. Like antibodies and TCRs, the KIRs exploit potentially mobile loops to mediate critical ligand contacts. In both the KIR2DL1/HLA-Cw4 and KIR2DL2/HLA-Cw3 structures, the in˚ 2 of solventteraction between the KIR and HLA-C buries approximately 1500 A ˚ 2 of buried accessible surface area, which is comparable to the 1700 to 1900 A surface in TCR/MHC complexes (99–101). Application of the algorithm of Lawrence & Colman (102) for quantitating shape complementarity in proteinprotein interfaces to the KIR2DL1/HLA-Cw4 and KIR2DL2/HLA-Cw3 complexes gives shape correlation statistics (Scs) of 0.71 and 0.69, respectively (51). (Sc = 1 for interfaces with perfect fits.) These values fall at the upper end of the ranges for TCR/MHC (Sc = 0.46–0.70) and antigen-antibody complexes (Sc = 0.64–0.68), indicating high shape complementarity. Charge complementarity due to one basic and six acidic residues on KIR2DL2, complemented by six basic residues on HLA-Cw3, results in the formation of four salt bridges (Glu21-Arg69, Glu106-Arg151, Asp135-Arg145, and Asp183Lys146) and eight hydrogen bonds between KIR2DL2 and HLA-Cw3. The crucial importance of these salt bridges for the KIR/HLA interaction was revealed by individually mutating each of the four KIR residues to alanine and measuring binding of the mutant KIRs to HLA-Cw3 by surface plasmon resonance. Mutation of Glu106 resulted in a sixfold lower affinity, and mutations of Asp135 and Asp183 each resulted in a 20-fold lower affinity for HLA-C. The sixteen KIR2DL2 residues that lie in the interface with HLA-Cw3 are all conserved in KIR2DL3, suggesting an identical mode of interaction between KIR2DL3 and HLA-Cw3. Remarkably, a comparison of the KIR2DL1/HLA-Cw4 and KIR2DL2/HLACw3 structures showed that many of the residues conserved in KIR2D and HLA-C mediate different interactions in the two complexes, apparently as a consequence of side chain rearrangements (51). Polar and charged residues at interfaces of both molecules form an extensive network of salt bridges and direct and water-mediated hydrogen bonds. Moreover, the interfaces exhibit strong charge complementarity: KIR2D has an electronegative binding surface rich in acidic residues that juxtaposes an electropositive surface on HLA-C in which the majority of contacting

8 Feb 2002

11:4

AR

AR152-26.tex

AR152-26.SGM

LaTeX2e(2001/05/10)

P1: IKH

NK CELL RECEPTORS

861

residues are basic. This type of highly charged interface, comprising a number of bridging salt links, is also observed in the Ly49A/H-2Dd complex (see below), and in the cell-cell adhesion complex between CD2 and CD58 (103). It may therefore be characteristic of receptor pairs requiring rapid dissociation and exchange to new partners for optimal cell signaling (NK receptors) or dynamic binding (CD2/CD58) (103).

Annu. Rev. Immunol. 2002.20:853-885. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

Structural Basis for Allelic Specificity The basis for the allelic specificity of KIR2DLs is apparent from the crystal structures of the KIR2DL1/HLA-Cw4 (51) and KIR2DL2/HLA-Cw3 (52) complexes. Eleven of 12 HLA-Cw3 residues in direct contact with KIR2DL2 are invariant in all HLA-C alleles, including HLA-Cw4. The only exception is Asn80, which mutagenesis studies have identified as critical in determining allotype specificity in this system. Of the 16 KIR2DL2 residues in the interface, all are conserved in KIR2DL3, and 14 are identical in KIR2DL1. The two residues that differ between KIR2DL2 and KIR2DL1 are at positions 44 and 70. In the KIR2DL2/HLA-Cw3 complex (50), Lys44 of KIR2DL2 forms a hydrogen bond with Asn80 of HLACw3 (Figure 3C). This hydrogen bond would be lost if Lys44 of KIR2DL2 were replaced by methionine (as in KIR2DL1), or if Asn80 of HLA-Cw3 were replaced by lysine (as in HLA-Cw2, 4, 5, 6, and 15). The KIR2DL1 receptor employs an alternative mechanism to achieve allelic specificity. In the KIR2DL1/HLA-Cw4 complex (51), the side chain of Lys80 of HLA-Cw4 is nestled in an electronegatively charged “specificity pocket” formed by a channel between the two KIR2DL1 domains. This pocket includes the polymorphic Met44 of the KIR that forms both polar and hydrophobic interactions with the side chain of Lys80 of HLA-Cw4 (Figure 3D). Additionally, Lys80 forms a hydrogen bond with Ser184 and a salt bridge with Glu187, both of which are conserved among KIRs. A substitution of Met44 with lysine found in KIR2DL2 would cause steric hindrance and charge repulsion with Lys80 of HLA-Cw4 leading to a loss of binding. Conversely, replacement of Lys80 of HLA-Cw4 with asparagine found in HLA-Cw3 and related alleles would result in a loss of both the hydrogen bond and the salt bridge, also leading to a loss of binding. Thus only the interaction of KIR2DL1 with HLA-Cw4 and related alleles would be energetically favored (51).

Structural Basis of Peptide Selectivity Although the overall docking mode of KIR2Ds onto MHC-I resembles that of TCRs, the contribution of the bound peptide to complex formation is markedly different in the two cases. Specific recognition of MHC/peptide by TCRs is largely dependent on peptide sequence, with several CDR loops of the TCR in intimate contact with the central portion of the antigenic peptide, at and around the P5 position (95–97, 99, 100, 104, 105). The KIR/HLA interaction is particularly

8 Feb 2002

11:5

Annu. Rev. Immunol. 2002.20:853-885. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

862

AR

AR152-26.tex

AR152-26.SGM

LaTeX2e(2001/05/10)

P1: IKH

NATARAJAN ET AL.

sensitive to substitutions at peptide positions 7 and 8 (84, 106, 107), a region that is structurally characterized by less extensive interactions (50, 51) than those seen in TCR/MHC structures. The KIR binding site is centered near residues P7 and P8 of the peptide that are near the C-terminus (Figure 2C and 2D). The KIR2DL1 surface around the P8Lys of the HLA-Cw4-bound peptide is overall electronegatively charged, and thus substitution of P8Lys with a negatively charged residue abolishes binding due to charge repulsion (51). In the analogous region of the KIR2DL2/HLA-Cw3 structure, the amide nitrogen of P8Ala is hydrogen bonded to Gln71 of KIR2DL2, consistent with preference for amino acids with small side chains like alanine or serine at the P8 position. For HLA-Cw3-bound peptides, larger amino acids such as lysine or tyrosine at P8 prevent binding to the KIR. However, substitutions at positions other than P8 are generally tolerated by KIR2Ds, in agreement with the complex structures. No significant changes in peptide or MHC conformation that could be attributed to receptor binding were noted from a comparison of the structures of free and KIR2DL1-bound HLA-Cw4 (51, 108). In particular, residue P7 is completely buried in the presence or absence of KIR2DL1. In the KIR2DL2/HLA-Cw3 complex, by contrast, the P7 side chain protrudes from the HLA-Cw3 peptide-binding cleft, suggesting that the binding of KIR2DL2 may induce a conformational change in the peptide that results in exposure of residue P7 to the NK receptor (50). However, confirmation of this hypothesis must await a structure determination of HLA-Cw3 in free form.

Insights into KIR Signaling While it is unknown how MHC-I binding to KIRs triggers the transmission of signals to the NK cell, two general mechanisms may be considered: (a) conformational changes in KIR structure upon ligand binding, and (b) formation of ligandinduced KIR oligomers. The finding that no major structural rearrangements occur in KIR2DL2 upon engagement of HLA-Cw3 effectively excludes conformational changes in the receptor as a mechanism for signal transduction. Although the angle between the D1 and D2 domains of KIR2DL1 differs by about 10◦ in the free and bound forms of the receptor, this probably simply reflects the need to adopt a particular interdomain orientation to assure specific contacts between the D1 binding loops and the α1 helix of HLA-Cw3, as well as between D2 and the α2 helix. A provocative feature of the KIR2DL2/HLA-Cw3 crystal is that the complex forms a regular oligomeric aggregate in the crystal lattice in which β-strands B and E of the KIR2DL2 D2 domain pack against the C-terminal portion of the α2 helix of HLA-Cw3 (50). This two-dimensional lattice may represent a physiological interaction at the interface between the NK cell and target cell that could promote KIR clustering on the NK cell surface, leading to receptor signaling. It should be noted, however, that no such arrangement of complex molecules was observed in the KIR2DL1/HLA-Cw4 crystal (51).

8 Feb 2002

11:7

AR

AR152-26.tex

AR152-26.SGM

LaTeX2e(2001/05/10)

P1: IKH

NK CELL RECEPTORS

863

Annu. Rev. Immunol. 2002.20:853-885. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

Although the structures of several KIRs and KIR/HLA complexes offer an understanding of the allelic specificity of the KIRs as well as the peptide preferences that they exhibit, the broader questions of the arrangement of the three domain KIR3D molecules and their mode of recognition of HLA-A and HLA-B ligands are largely unanswered. Whether the additional domain contributes to the potential for oligomerization or to additional restraints on MHC or peptide specificity remains to be determined. Further structural studies promise to enhance our knowledge on these matters.

LEUKOCYTE IMMUNOGLOBULIN-LIKE RECEPTOR-1 Leukocyte immunoglobulin-like receptor-1 (LIR-1; also known as ILT2 or CD85h; see Table 1) is an inhibitory receptor, whose family is widely distributed on monocytes, B cells, dendritic cells, and some NK and T cells. LIR-1 recognizes a broad range of MHC-I molecules, as well as UL18, an MHC-I homolog encoded by human cytomegalovirus (109–115). The LIR/ILT family consists of about 13 related members, Ig-like in structure, analogous to KIRs (Table 1). LIR proteins contain various numbers of extracellular Ig-like domains: LIR-1, -2, -3, -4, -6a, -7, and -8 have four such domains, whereas LIR-5 and -6b have only two. and the presence of ITIM sequences in the cytoplasmic regions of LIR-1, -2, -3, -5, and -8 suggests that these LIRs may act as inhibitory receptors in a manner analogous to KIRs (116). The crystal structure of domains 1 and 2 (D1 and D2) of LIR-1, which contain the ligand-binding site, has been determined (52). As shown in Figure 1B, the two tandem immunoglobulin domains form a bent structure with an acute interdomain angle, similar to KIR2DL1-3. Each domain is composed of two anti-parallel βsheets arranged in a KIR-like topology. LIR-1 displays the same strand switch observed in the KIR2D structures (47–49). Unlike KIR2D, however, LIR-1 includes several 310 helical regions interspersed in the primarily β structure. The C0 β-strand seen in the KIR structures is replaced by a 310 helix in D1 of LIR-1. Another significant difference is the presence of a polyproline II type helix in the F-G loop of both LIR-1 domains. The interdomain interface of LIR-1 is formed by contacts between the E-F loop and strand G regions of D1 with the F-G loop of D2, and the hinge region. This interface is dominated by hydrophobic interactions, as in the KIR2D structures. Despite the striking structural similarities between LIR-1 and KIR2D, the two receptors appear to have different ligand-binding sites. While the crystal structure of a LIR-1/ligand complex has not been reported, mutagenesis and direct binding experiments have localized the UL18 binding site of this receptor, which may correspond to that for MHC-I, to the A0 CC0 FG face of the D1 domain (52) (Figure 1B). This site is distant from the elbow region between D1 and D2 that contacts MHC-I in the KIR2DL1/HLA-Cw4 (51) and KIR2DL2/HLA-Cw3 complexes (50), illustrating how members of the same protein family may employ completely different surfaces for recognizing structurally similar ligands.

4 Feb 2002

12:4

864

AR

AR152-26.tex

AR152-26.SGM

LaTeX2e(2001/05/10)

P1: IKH

NATARAJAN ET AL.

Annu. Rev. Immunol. 2002.20:853-885. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

C-TYPE LECTIN-LIKE RECEPTORS Although the Ig-superfamily has provided the major participants in MHC-I recognition by NK cells in the human and other primates, the C-type lectin-like family constitutes the primary MHC-I-recognizing NK receptors of the mouse. However, the human and the mouse share other members of the C-type lectin-like family that recognize MHC-I related molecules. Members of the C-type lectin-like family of NK receptors (Ly49A through W, NKR-P1, CD94/NKG2A/B, -C or -E, NKG2D, CD69) are homodimeric (Ly49A through W, NKR-P1, NKG2D, CD69) or heterodimeric (CD94/NKG2A, -B, -C, or -E) type II (intracellular amino terminus) transmembrane glycoproteins. Each subunit is composed of an extracellular C-type lectin-like domain (CTLD) connected by a stalk region of 25-75 residues to transmembrane and cytoplasmic domains. In addition to the C-type lectin-like domains of NK receptors (known as NK receptor domains—NKDs), the CTLD superfamily of protein modules, as defined previously (2, 3), includes the carbohydrate-recognition domains (CRDs) of C-type (or Ca2+-dependent) lectins, such as mannose-binding protein (MBP) and the selectins. The structural similarity between NKDs and CRDs, which constitute separate families within the CTLD superfamily, probably reflects evolution from a common ancestral domain but does not imply functional similarity. Thus, as discussed below, NKDs appear to be primarily protein-binding, rather than sugar-binding, molecules. Protein-binding CTLDs are also found in coagulation factor IX/X-binding protein (117, 118) and in CD23 (119), which recognizes IgE. Other CTLDs may bind inorganic surfaces, such as those of ice (type II antifreeze glycoprotein) (120, 121) and calcium carbonate crystals (lithostathine) (122). It is noteworthy that other members of the CTLD superfamily, in addition to the C-type lectin-like NK receptors, are critically involved in mediating innate immune responses. These include the collectins (MBP, pulmonary surfactant proteins, bovine conglutinin) that neutralize a broad range of bacterial and fungal pathogens through aggregation or complement activation (123), and the macrophage mannose receptor that has been implicated in the phagocytosis of microbial pathogens such as Candida albicans and Mycobacterium tuberculosis (124, 125). The following description of the function and structure of NK receptors of the C-type lectin-like family first gives a brief overview of the biology of the Ly49 family, CD94/NKG2, and NKG2D. Then we systematically compare the structures of the monomeric subunits of these molecules, their mode of dimerization, and the mechanism of ligand recognition as revealed by structures of ligand/receptor complexes.

The Ly49 Family Ly49 is a multigene family of homodimeric C-type lectin-like receptors expressed on mouse NK cells and also on a subset of cytotoxic T cells (126). While the KIRs serve as the major functional NK receptor family recognizing MHC-I molecules

4 Feb 2002

12:4

AR

AR152-26.tex

AR152-26.SGM

LaTeX2e(2001/05/10)

P1: IKH

Annu. Rev. Immunol. 2002.20:853-885. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

NK CELL RECEPTORS

865

in the human, the mouse utilizes members of the Ly49 family for the analogous function, that is, recognition of MHC-I molecules, H-2K and H-2D. cDNA cloning has identified 11 functional Ly49 genes in the C57BL/6 strain (127). Additional Ly49s not found in C57BL/6 have been found in the 129/J strain (128), suggesting strain variations in the Ly49 repertoire as the result of a rapid rate of evolution. The repertoire of Ly49s expressed by an individual NK cell includes both activating and inhibitory types, functions transduced through either cytoplasmic ITIMs or kinase-associated adapter molecules such as DAP10 or DAP12 (15, 21). Thus, Ly49 receptors regulate NK function through a complex and as yet poorly understood mechanism involving a dynamic balance of activating and inhibitory signals received through receptors that recognize MHC-I. While the mouse and other rodents rely on members of the Ly49 family as their major recognition receptors for modulating MHC-I recognition, the human retains only a single, apparently nonfunctional remnant of the Ly49 family, Ly49L (129, 130). An intermediate solution has been provided in the baboon, where functional transcripts of both an Ly49L equivalent and KIRs have been identified (131). One of the best-studied MHC/Ly49 interactions is that between the inhibitory receptor Ly49A and its H-2Dd ligand. Cytotoxicity of Ly49A+ NK cells is inhibited when targets express the H-2Dd molecule, either naturally or by transfection (11). As shown in experiments using an exogenously loadable, H-2Dd-transfected, TAPdeficient cell line as targets, recognition of H-2Dd by Ly49A+ effectors in a killing assay was not influenced by the amino acid sequence of the bound peptide (132). In vivo, the level of Ly49A expression per cell is decreased in mice expressing H-2Dd but not in strains that do not express ligand (133). In order to investigate in greater detail the molecular basis of the specificity of the H-2Dd/Ly49A interaction, a recombinant soluble version of Ly49A encoding the entire extracellular region (designated Ly49A-EC) was prepared by bacterial expression and in vitro folding and purification. The binding of Ly49-EC to similarly expressed H-2Dd was then examined in in vitro binding assays (134) and structurally by high resolution X-ray crystallographic analyses of the Ly49A/H-2Dd complex (57). The in vitro folded molecule displays epitopes for several Ly49A-specific monoclonal antibodies and, as evaluated by surface plasmon resonance, binds to bacterially expressed and folded H-2Dd. Thus, although related to C-type lectins, Ly49A does not appear to require glycans for recognition. Buffers used in Ly49A refolding, purification, and binding studies did not contain Ca2+, and this lack of a Ca2+ requirement is in agreement with the lack of a Ca2+ binding motif in Ly49A based on amino acid sequence comparisons with mannose binding protein (3, 57). Irrespective of the peptide used to assemble H-2Dd, the KD were in the range of 6 µM to 20 µM, a value similar to those reported for most TCR/MHC interactions (135, 136). The Ly49A-EC/ H-2Dd interaction, studied at 25◦ C, was governed by kon and koff of ∼2000 M−1 sec−1 and 0.03 sec−1, respectively. Ly49A-EC failed to compete with an H-2Dd-restricted, P18-I10 specific, soluble single-chain TCR (137) for binding to H-2Dd/P18-I10, indicating that Ly49A-EC and TCR bind to

4 Feb 2002

12:4

866

AR

AR152-26.tex

AR152-26.SGM

LaTeX2e(2001/05/10)

P1: IKH

NATARAJAN ET AL.

distinct, non-overlapping sites on H-2Dd, an observation that was confirmed by the structure of the complex (see below).

Annu. Rev. Immunol. 2002.20:853-885. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

The CD94/NKG2 Family The CD94/NKG2 receptors are lectin-like, disulfide-linked heterodimers expressed as type II membrane proteins on human and rodent NK cells and a subpopulation of T cells. Genes encoding CD94 and NKG2 are located within the NK gene complex of humans and rodents. The CD94 subunit of the receptor is invariant and is encoded by a single gene (138). NKG2 on the other hand constitutes a multigenic family of at least five proteins, designated NKG2A (and its splice variant NKG2B), NKG2C, NKG2D, and NKG2E. The extracellular, ligand-binding domains of NKG2A/B, -C, and -E share a high degree of homology. The cytoplasmic domains are either long (NKG2A/B) or short (NKG2C and -E), corresponding to inhibitory or activating CD94/NKG2 isoforms, respectively. Inhibitory isoforms have a pair of ITIM motifs in their cytoplasmic tails that upon phosphorylation recruit SHP-1 and SHP-2 phosphatases to block cellular signaling (139). The activating isoforms CD94/NKG2C and -E do not directly signal but instead associate with a short, dimeric ITAM-containing signaling molecule KARAP12/DAP12 (140, 141). Although the NKG2D gene is located adjacent to other NKG2s within the NK complex, its encoded protein shows little homology to other NKG2s and does not pair with CD94. The expression and function of NKG2D is discussed separately below. Ligands for CD94/NKG2A are the nonclassical MHC-I molecules HLA-E and Qa-1, in humans (142–144) and mice (145), respectively. HLA-E and Qa-1 selectively bind peptides derived from the leader sequences of MHC-I heavy chains (146–148), and thus proper expression of HLA-E or Qa-1 is an indicator of normal expression of MHC-I molecules. Tetrameric complexes of HLA-E specifically stained transfectants expressing CD94/NKG2A , -B, or -C in human (149) and Qa-1 tetramers specifically stained mouse CD94/NKG2A transfectants (145). HLA-E expression protected otherwise susceptible target cells from lysis by CD94/NKG2A expressing NK clones (148–150). Reciprocally, fluorescently tagged soluble CD94/NKG2A produced in an insect expression system specifically stained HLAE transfected RMA-S cells but not those transfected with HLA-A, -B, or -C alleles (151). Direct binding between soluble CD94/NKG2 receptors and soluble HLA-E ligands was demonstrated by surface plasmon resonance techniques using bacterially expressed and in vitro folded molecules (152). Comparison of the HLA-E binding parameters of the inhibitory receptor CD94/NKG2A and its activating counterpart CD94/NKG2C revealed extremely rapid association and dissociation rates for both interactions. Similar to the observations on the KIR2D interaction with HLA-C (75), the interaction with the inhibitory receptor is of higher affinity than with the activating isoform. The molecular basis of the difference in affinities, although not yet illuminated by a crystal structure of the complex, may be due to a lysine/glutamic acid dimorphism at position 197.

4 Feb 2002

12:4

AR

AR152-26.tex

AR152-26.SGM

LaTeX2e(2001/05/10)

P1: IKH

NK CELL RECEPTORS

867

Annu. Rev. Immunol. 2002.20:853-885. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

NKG2D NKG2D is a member of the C-type lectin-like family of NK receptors that is only distantly related to NKG2A, -B, -C, or -E (20%–30% sequence identity). This homodimeric activating receptor is expressed on NK cells, CD8+ αβ T cells, and γ δ T cells (153). NKG2D is associated with the DAP10 adapter molecule, through which it triggers NK cell–mediated lysis of certain tumor cells (65). In contrast to the heterodimeric CD94/NKG2 receptors that bind HLA-E, human NKG2D recognizes MICA and MICB, MHC-I homologs composed of α1, α2, and α3 domains. These NKG2D ligands lack β2m and do not bind peptides. Unlike classical MHC-I proteins, MICA and MICB are minimally expressed on normal tissues but are upregulated in stressed cells and overexpressed by epithelial tumors (154–156). In addition to MICA and MICB, a group of proteins designated ULBPs, which bind cytomegalovirus glycoprotein UL16, have been identified as ligands for human NKG2D (157). ULBPs are composed of an α1/α2 platform domain, lack an α3 domain and β2m, and are anchored to the membrane by a glycophosphatidylinositol (GPI) linkage. It has been proposed that UL16 acts as a decoy receptor for NKG2D ligands, facilitating viral evasion of the immune system. Rodent homologs of MICA and MICB have not been identified, but other molecules with weak homology to MHC-I, including RAE-I and H-60, serve as ligands for mouse NKG2D (158, 159). While little is known about their function, RAE-1 and H-60, like MICA and MICB, are frequently expressed on tumor cells but not normal tissues. Like ULBPs, RAE-1 and H-60 contain only the homologous α1 and α2 domains of MHC-I, are GPI-linked, and do not bind peptides. Recent evidence suggests that NKG2D may serve as an effective receptor for a tumor surveillance function of NK cells, exploiting the upregulation of Rae-1 and/or H-60 on tumor targets that provide a ligand for this activating receptor (160). The crystal structures of human NKG2D bound to MICA (58) and ULBP (S Radaev, B Rostro, AG Brooks, M Colonna, PD Sun, submitted) and of murine NKG2D in complex with RAE-1 (P Li, G McDermott, RK Strong, submitted) have been determined, revealing the molecular basis for the remarkable promiscuity of the NKG2D binding site.

The NKD Fold The CTLDs of NK receptors of known three-dimensional structure [currently Ly49A (57), Ly49I (N Dimasi, MW Sawicki, LN Reineck, Y Li, K Natarajan, DH Margulies, RA Mariuzza, unpublished data), CD94 (53), NKG2D (56, 58) and CD69 (54, 55)] exhibit a highly conserved folding pattern consisting of two α-helices (α1 and α2), except in CD94 where the first α-helix is replaced by an extended loop, and two anti-parallel β-sheets formed by β-strands β0, β1, β5; β2, β20 , β3, β4 (Figures 2B and 4). With the exception of strand β0, a characteristic of long-form CTLDs such as tetranectin (161, 162) and lithostathine (122), this configuration is observed in all CTLDs whose structures have been determined to

8 Feb 2002

11:9

Annu. Rev. Immunol. 2002.20:853-885. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

868

AR

AR152-26.tex

AR152-26.SGM

LaTeX2e(2001/05/10)

P1: IKH

NATARAJAN ET AL.

date. The core regions—composed of strands β1, β2, β20 , β3, β4, and β5, and helix α1—of these NKDs are virtually superimposable. However, considerable variation is seen in the sequence and conformation of the loops connecting the β-strands. Like the complementarity-determining loops (CDRs) of antibodies and TCRs, certain of these are responsible for conferring ligand-binding specificity to the NKD. Ly49A, Ly49I, CD94, NKG2D, and CD69 loops 1 and 2 are similar structurally, whereas the conformations of loops 3 and 5 are quite different. This is consistent with the finding that loop 3 constitutes a major portion of the crystallographically visualized (Ly49A, NKG2D) or presumed (Ly49I, CD94, CD69) ligand-binding site of these NKDs. The conformation of loop 5 seems to be influenced by the orientation of the α2 helix. There are three conserved disulfide bonds in the NKD structures (Figure 4). Those joining helix α1 and β5, and β3 and the loop following β4, are invariant in C-type animal lectins. The third conserved disulfide links strands β0 and β1, and it is also present in other long-form CTLDs. In the crystal structures of the Ly49A/H-2Dd (57), NKG2D/MICA (58), NKG2D/RAE-1β (P Li, G McDermott, RK Strong, submitted), and NKG2D/ ULBP3 (S Radaev, B Rostro, AG Brooks, M Colonna, PD Sun, submitted) complexes, the MHC-binding site of the NKDs extends from loop 3 to β-strand 5; this region corresponds to the sugar-binding site of MBP-A (Figure 4). Although the ligand-binding sites of other NKDs (CD94/NKG2A, CD69, NKR-P1) remain to be defined, it appears likely that the same general region will be implicated. This region is between five and nine residues shorter in the NKDs than in the MBP-A CRD (3), which has an additional loop (loop 4) that contributes to Ca2+ binding (Figure 4). In addition, the NKDs do not retain residues involved in coordinating Ca2+ at either of two sites in MBP-A; no bound Ca2+ is observed in the crystal structures of Ly49A (57), Ly49I (N Dimasi, MW Sawicki, LN Reineck, Y Li, K Natarajan, DH Margulies, RA Mariuzza, unpublished data), CD94 (53), NKG2D (56, 58) (S Radaev, B Rostro, AG Brooks, M Colonna, PD Sun, submitted), or CD69 (54, 55). Thus, if these NKDs retain sugar-binding capacity, it is not mediated through Ca2+ ions, as in C-type lectins such as MBP-A and the selectins. The known MHC-binding surfaces of Ly49A and NKG2D, and the putative ligandbinding sites of Ly49I and CD94, are relatively flat, consistent with their function as protein-binding receptors.

The NKD Dimer At the cell surface, C-type lectin-like NK receptors exist as disulfide-linked homodimers (Ly49, NKG2D, CD69), or as heterodimers in the case of the CD94/ NKG2A/B, -C or -E (163). In the structures of Ly49A, Ly49I, NKG2D, CD69, and the CD94 crystallographic dimer, the interface between the NKD monomers is relatively flat and predominantly hydrophobic. The subunits associate through their first β-strand, β0, creating one extended six-stranded anti-parallel β-sheet (Figure 5). In addition, the C-terminal ends of helix α2 pack against one another, except in the case of CD94, where the α2 helix is replaced by a loop,

8 Feb 2002

11:10

AR

AR152-26.tex

AR152-26.SGM

LaTeX2e(2001/05/10)

P1: IKH

Annu. Rev. Immunol. 2002.20:853-885. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

NK CELL RECEPTORS

869

and in Ly49I, where the α2 helix is not involved in the dimer interface. However, since the biological form of CD94 is a heterodimer with NKG2A/B, -C or -E, it is unclear whether the absence of this helix in the CD94 structure represents an authentic feature of the CD94/NKG2 heterodimer. Thus far, this mode of dimerization appears unique to NKDs. Other CTLDs known to form physiological dimers, such as coagulation factor IX and tunicate lectin, associate differently. Coagulation factor IX dimerizes through interactions between the loop regions connecting strands β2 and β3 (117, 118), whereas tunicate lectin dimerizes via β-strand 1 and extensive interactions along the entire length of the α2 helix (164). The region of the dimer interface formed by the anti-parallel β0 strands is similar among the members of this structural family, with at least two main chain-main chain hydrogen bonds linking the strands. These strands contain a conserved W-XX-Y/H motif in which the second residue is hydrophobic and packs against itself. In CD69 and NKG2D, flanking charged residues form salt bridges across the interface that may provide additional stabilization. The juxtaposition of the α2 helices in the Ly49A, CD69, and NKG2D NKDs reveals greater structural variability than the β0 portion, but in all three examples comprises a tight hydrophobic core primarily composed of aromatic residues. These two components of the NKD interface (i.e., the β0 strands and the α2 helix contacts) together bury a surface area in the range ˚ 2 (NKG2D). ˚ 2 (Ly49A) to 2200 A of 930 A The structure of Ly49I reveals significant variability in dimerization mode within the Ly49 family (N Dimasi, MW Sawicki, LN Reineck, Y Li, K Natarajan, DH Margulies, RA Mariuzza, unpublished data). Whereas the β0 interface of Ly49I is similar to those of the Ly49A, NKG2D, and CD69 NKDs, the Ly49I monomers are further linked by a β-hairpin formed by the C-terminal half of strand β0 and the N-terminal end of β1. Conversely, the α2 helix of Ly49I does not contribute to the dimer interface, opening up the Ly49I dimer relative to Ly49A, NKG2D, or CD69 (Figure 5). As a result, the putative MHC-binding surfaces of the Ly49I dimer are somewhat more separated spatially than the corresponding surfaces of Ly49A, and much more so than those of the NKG2D dimer. These structural differences may reflect the fundamentally different ways in which Ly49 and NKG2D receptors recognize their respective ligands: whereas the single MICA binding site of NKG2D is formed by the precise juxtaposition of two monomers, each Ly49 monomer appears to contain an independent binding site for MHC-I, as discussed below. Hence, the structural constraints on dimerization geometry may be relatively relaxed within the Ly49 family. Presently, no Ly49 family member has had its structure solved crystallographically in both the free and liganded forms. A complete appreciation of the changes that might take place on ligand binding awaits such comparison.

Structure of the Ly49A/H-2Dd Complex The Ly49A homodimer interacts with two distinct sites on H-2Dd, designated Site 1 and Site 2 (57) (Figure 6). In agreement with experiments showing lack of competition between Ly49A and TCR binding to H-2Dd (134),

4 Feb 2002

12:4

Annu. Rev. Immunol. 2002.20:853-885. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

870

AR

AR152-26.tex

AR152-26.SGM

LaTeX2e(2001/05/10)

P1: IKH

NATARAJAN ET AL.

neither site overlaps the TCR binding site. The interface at Site 1 is at one end of the peptide binding groove and spans the N terminus of the α1 helix and C terminus of the α2 helix of H-2Dd. This site also includes the partially conserved N-linked ˚ 2, glycosylation site at position 176. The total buried surface area is about 1000 A 2 ˚ which is at the lower end of the average of 1600 ± 400 A for protein-protein recognition sites (165) and is similar in size to the heterophilic adhesion complex between CD2 and CD58 (103). At Site 1 there is a precise match of the topology of the interacting surfaces as reflected by a high value for the shape correlation statistic (102) (Sc index = 0.78, comparable to that of protease-protease inhibitor interactions that have an Sc index = 0.70–0.76). The interface at Site 2 is a broad cavity underneath the peptide-binding platform of the MHC in a region that overlaps the CD8 binding site (166, 167). Here the Ly49A dimer makes contacts with α1/α2, α3, and β2m domains of the MHC, which contribute 60%, 15%, and 25% of the buried surface, respectively. The ˚ 2) total buried surface area of Site 2 is much larger than that of Site 1 (>3,300 A and shows poor shape complementarity (Sc value = 0.54 without bound water) similar to that reported for TCR/MHC-peptide complexes (Sc values ranging from 0.46 to 0.63) (99). The contribution of Ly49A subunits of the homodimer, Ly49A-1 and Ly49A-2, to the interaction at Site 2 is unequal, with 79% of the buried surface contributed by Ly49A-2 alone. Ly49A-1 makes a minor contribution to this site and contacts the N terminus of the α2 helix and C terminus of the α helix of the MHC in a location close to the widely conserved N-linked glycosylation site at Asn86. The close proximity of both N-linked glycosylation sites of H-2Dd, Asn86 and Asn176, to the Ly49A binding sites raises the possibility that N-linked oligosaccharides on H-2Dd may modulate the interaction with Ly49A (168).

The Site1/Site 2 Conundrum and Its Resolution The crystallographic visualization of two potential binding sites on the MHC-I molecule H-2Dd posed a novel set of questions that have been the subject of intense investigation in several laboratories including our own: (a) Were both Sites 1 and 2 of biological significance? That is, are they functional in specific binding of the MHC to the NK receptor? (b) What residues of the NK receptor and the MHC molecule are critical to the interaction? (c) Can the primary focus of the interaction explain the MHC specificity of Ly49A? (d ) Can it explain the peptide preference that some alleles of Ly49 seem to exhibit? Our initial interpretation of the structural evidence in the context of data concerning the allelic preferences of the Ly49A binding and a variety of functional studies was based on several points: (a) the degree of shape complementarity at each of the two interfaces; (b) the location of polymorphic residues of MHC-I molecules that might explain specificity of recognition by Ly49 family members; (c) the likely location of the binding site relative to whether the H-2Dd/Ly49A interaction was primarily between molecules on different cells, regulating recognition signals between NK and target, or on the same cell, regulating surface expression

8 Feb 2002

11:12

AR

AR152-26.tex

AR152-26.SGM

LaTeX2e(2001/05/10)

P1: IKH

Annu. Rev. Immunol. 2002.20:853-885. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

NK CELL RECEPTORS

871

of Ly49 as a function of H-2Dd level; (d ) which site could best explain the peptide dependent, but peptide non-specific recognition; and (e) whether the complete lack of functional recognition of a single chain β2m/H-2Dd chimeric molecule by Ly49A (169) could be explained by the stucture. Because of the very precise interface at Site 1 with high shape complementarity, the lack of polymorphic residues at Site 2, and the accessibility of Site 1 for an interaction between Ly49A on an NK cell and H-2Dd on a target, we initially favored Site 1 as the major site of interaction. We considered that Site 2 might play a secondary role in the regulation of expression of Ly49A on H-2Dd positive cells. Subsequently, using recombinant Ly49A labeled with biotin as a cell staining reagent, we tested one Site 1 mutant, H-2Dd I52M, and found a small effect on Ly49A binding (170). More recently, Matsumoto and colleagues have examined a number of H-2Dd mutants both in assays of Ly49A function and in binding analysis using a recombinant Ly49A reagent similar to that used by Chung et al. (170), and observed that several residues of H-2Dd involved in Site 2 contacts made major contributions to the interaction (171). In a comprehensive evaluation of mutants of both Ly49A and of H-2Dd, we have identified residues of Ly49A, H-2Dd, and β2m that are critical to the Ly49A/H-2Dd interaction (171a). These mutations pinpoint crucial residues of contact in the interface at Site 2, implicating this large surface area as the major determinant of Ly49A/H-2Dd functional interaction. Allelic and peptide preference as well as contributions from α2m can be explained by contacts involving H-2Dd residue Asp122 and α2m residues Gln29 and Lys58. The H-2Dd residue Asp122 is the focus of interactions from Ly49A residues 236, 238, and 239. α2m residue Gln29 makes contacts with Ly49A residues 247, 248, and 249; and α2m residue Lys58 contacts 229, 239, 241, and 242 of Ly49A. In addition to analysis of Ly49A/H-2Dd contacts, we studied a number of residues involved in the Ly49A homodimer interface. Individual mutations of several of these significantly reduce binding to H-2Dd, indicating a need for proper homodimerization to permit binding to the MHC-I molecule. The general conclusion based on both structural and mutational analysis is that the Ly49A homodimer makes its functional binding interaction at Site 2 (171a). Allelic specificity is likely to be controlled by polymorphic residues at the perimeter of this binding site, and it may in part be dependent on a number of water molecules that organize contacts in the interface. Since no structural information concerning the stem and neck regions of Ly49A is yet available, and analysis of chimeric molecules has suggested the importance of the neck region in molecular specificity (172), the structural contribution of the stem and neck to binding, specificity, and signaling remains to be determined.

Structure of the Human NKG2D/MICA Complex MICA, whose crystal structure in free and NKG2D-bound forms has been determined (58, 173), is composed of two structural domains, an α1/α2 platform domain and a C-type Ig-like α3 domain (Figure 7A). The α1/α2 platform comprises four

8 Feb 2002

11:12

Annu. Rev. Immunol. 2002.20:853-885. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

872

AR

AR152-26.tex

AR152-26.SGM

LaTeX2e(2001/05/10)

P1: IKH

NATARAJAN ET AL.

distinct α-helices, designated 1 and 2 in α1 and 1 and 2b in α2, arranged on an eight-stranded anti-parallel β-sheet. These helices correspond to the α1 and α2 helices that define the peptide-binding groove in classical MHC-I molecules. In the crystal structure of the human NKG2D/MICA complex (Figure 7A), the NKG2D homodimer binds orthogonally to the MICA platform in a manner resembling the docking mode of TCRs onto MHC-I, such that each NKG2D monomer mainly contacts either the α1 or α2 domain of the MICA ligand through the α-helical elements. This recognition of an asymmetric ligand by a symmetric receptor is mediated by very similar surfaces on the NKGD2 monomers that specifically interact with two distinct surfaces on MICA. Thus, 7 of the 11 contact residues contributed by each NKG2D monomer are common to both MICA binding surfaces, although six are engaged in different interactions at the two interfaces. In other cases of homodimeric receptors binding to monomeric ligands, by contrast, one subunit of the symmetric receptor has been found to dominate the interaction with the asymmetric ligand. Examples include the interaction of Ly49A with H-2Dd, discussed above, or the binding of CD8αα to HLA-A2 (166) or H-2Kb (167). ˚ 2 of The interaction between NKG2D and MICA buries approximately 2200 A solvent-accessible surface area, which is somewhat greater than the buried surface in TCR/MHC (99–101) or KIR2D/MHC (50, 51) complexes, but significantly less ˚ 2) at the Site 2 interface in the Ly49A/H-2Dd comthan the buried surface (3300 A plex (57). The NKG2D/MICA interface, like the KIR2D/HLA-C and Ly49A/H2Dd Site 2 interfaces, is characterized by a high degree of shape complementarity (Sc = 0.72). In contrast to these other interfaces, however, which are highly hydrophilic and dominated by polar and charged interactions, the NKG2D/MICA interface comprises a mixture of hydrophobic, polar, and charged interactions that is more typical of known protein-protein interfaces (165).

Conformational Changes in MICA upon NKG2D Binding In the crystal structure of free MICA (173), ten residues in the central portion of the α2 helix were observed to be disordered and presumed to constitute an extended, conformationally flexible loop. In the NKG2D/MICA complex, however, these residues are ordered into two additional turns of helix, thereby restoring the canonical α2 helical element of classical MHC-I proteins (58). The ordering of this loop, which appears to result from direct contacts with NKG2, creates a small cavity corresponding in location to the peptide-binding groove of MHC-I molecules. However, there is no indication from the crystal structure of a peptide or nonpeptide ligand bound in this cavity. The most pronounced difference between MICA in its bound and unbound forms is found in the relative orientation of the α1/α2 and α3 domains. In the isolated MICA structure, these domains adopt an extended arrangement that differs dramatically from the norm for classical MHC-I proteins. By contrast, the interdomain orientation of MICA in the NKG2D/MICA complex resembles that

8 Feb 2002

11:14

AR

AR152-26.tex

AR152-26.SGM

LaTeX2e(2001/05/10)

P1: IKH

NK CELL RECEPTORS

873

in MHC-I structures. It seems unlikely, however, that the difference in relative orientation of the platform and α3 domains in free versus liganded MICA is a consequence of NKG2D binding; rather, it probably simply reflects the intrinsic flexibility of MICA due to the lack of interdomain contacts and the absence of the light chain subunit, β2m.

Annu. Rev. Immunol. 2002.20:853-885. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

Structure of the Mouse NKG2D/RAE-1 Complex In the crystal structure of unbound RAE-1 (P Li, G McDermott, RK Strong, submitted) the two α-helices of the RAE-1 platform domain are covalently linked by a noncanonical disulfide bond, resulting in the loss of any vestige of a peptidebinding groove. In the structure of RAE-1 bound to mouse NKG2D (P Li, G McDermott, RK Strong, submitted), the NKG2D homodimer binds diagonally to the RAE-1 platform domain in a manner resembling the interaction of human NKG2D and MICA (Figure 7B). As in the human NKG2D/MICA complex, each mouse NKG2D monomer appears to make roughly similar contributions to ligand recognition. However, in contrast to the human NKG2D/MICA interaction, where the α2 helix of MICA undergoes a large structural rearrangement upon receptor binding, no significant conformational changes were observed in RAE-1 (or NKG2D) following complex formation. A detailed analysis of NKG2D/RAE-1 ˚ of the structure of the interactions is precluded by the limited resolution (3.5 A) complex. It is nevertheless apparent that, while most of the residues of mouse NKG2D in contact with RAE-1 are conserved in human NKG2D, they interact with a different set of ligand residues in ways different from the human NKG2D/MICA complex. Thus, NKG2D has evolved a highly conserved binding surface capable of promiscuous recognition of several nonclassical MHC-I ligands.

Structure of the Human NKG2D/ULBP Complex The crystal structure of human NKG2D in complex with ULBP (S. Radaev, B. Rostro, AG Brooks, P Sun, submitted) reveals that the NKG2D homodimer binds orthogonally across the α-helices of the α1 and α2 domains of ULBP (Figure 7C), similar to the interaction of NKG2D with MICA (Figure 7A) or RAE-1 (Figure 7B). The potential peptide-binding groove of ULBP is completely occluded by the side chains of hydrophobic residues from the opposing helices. Like the NKG2D/MICA complex (58), the NKG2D/ULPB complex is stabilized mainly by hydrophobic interactions and hydrogen bonds, with fewer salt bridges and less charge complementarity than in the highly hydrophilic KIR2D/HLA-C (50, 51) and Ly49A/H-2Dd (57) interfaces. A comparison of crystal structures of free (56) and ULBP-bound (S Radaev, B Rostro, AG Brooks, P Sun, submitted) mouse NKG2D indicates that the receptor undergoes conformational adjustments upon complex formation that are characteristic of an induced-fit recognition mechanism. In particular, the relative orientation of the two subunits in the bound NKG2D homodimer differs from that

8 Feb 2002

11:17

874

AR

AR152-26.tex

AR152-26.SGM

LaTeX2e(2001/05/10)

P1: IKH

NATARAJAN ET AL.

Annu. Rev. Immunol. 2002.20:853-885. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

in the free form by a small, but significant, rotation of 5◦ . Furthermore, differences in the conformation of one of the ligand-binding loops of NKG2D (L3 in Figure 4) are observed: In one of the monomers, this loop exhibits the same conformation as in the free receptor but has undergone a main-chain rearrangement in the other monomer. By recruiting different NKG2D residues to the interface and enabling the same residues to adopt different conformations in the two receptor subunits, this structural plasticity enables NKG2D to bind diverse ligand surfaces while maintaining the same overall docking orientation (Figure 7).

The CD94/NKG2A Dimer—Predictions To date, the only known structures of C-type lectin-like NK receptors in complex with MHC and MHC-like ligands are of those NK receptors that are homodimeric (Ly49A and NKG2D). Although the crystal structure of a nonphysiological CD94 homodimer has been reported, no structure of a CD94/NKG2 receptor either free or in complex with ligand has been determined. However, plausible models of the complex have been constructed (58) and (S Radaev, B Rostro, AG Brooks, P Sun, submitted) by superimposing: (a) CD94 (53) onto the NKG2D monomer that primarily contacts the α1 domain of MICA in the NKG2D/MICA complex (58), or that contacts the α1 domain of ULBP in the NKG2D/ULBP complex (S Radaev, B Rostro, AG Brooks, P Sun, submitted) and (b) the α1/α2 platform of HLA-E (174) onto that of MICA or ULBP. In these models, CD94 is positioned over the α1 helix of HLA-E and NKG2 over the α2, with the NKG2D subunit of the heterodimer in possible contact with the self-peptide bound to HLA-E. This model is consistent with binding studies that indicate that peptide position P8 bound to HLA-E interacts with the CD94/NKG2A heterodimer (152) and that peptide positions P4, P5, and P8 in the Qa-1 complex directly interact with murine CD94/NKG2A (175).

SUMMARY AND CONCLUSIONS Understanding the molecular interactions between receptors on NK cells and their MHC-I and MHC-I-like ligands provides not only insight into the mechanisms by which cells of the innate immune system discriminate normal from pathologically dysregulated cells, it offers an appreciation of the rapid and abundant evolution that accompanies this broad family of receptors. In the human, the KIRs of the Ig superfamily have taken hold for now, and in the mouse an analogous set of functions have been undertaken by the Ly49 family of the C-type lectin-like superfamily. Both of these structural templates, the Ig and the C-type lectins, seem to have solved the requirements for molecular stability in a tight core and molecular adaptability in a set of potentially variable and flexible loops in different ways. The robust structure of the MHC-I scaffold is also revealed, as different NK receptors,

8 Feb 2002

11:18

AR

AR152-26.tex

AR152-26.SGM

LaTeX2e(2001/05/10)

P1: IKH

NK CELL RECEPTORS

875

Annu. Rev. Immunol. 2002.20:853-885. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

as well as TCR, clearly interact with several distinct faces of the molecule (Figure 8). The molecular economy by which a single class of peptide receptor molecules and their structural relatives can interact with a wide spectrum of immune cell receptors is one profound example of the opportunities for specific recognition in a host of biological systems. Other regulatory molecules that will soon be better understood in both functional and structural contexts, such as the paired immunoglobulin-like family (176, 177), will certainly reflect the functional and structural paradigms so far elucidated by study of the NK receptors. ACKNOWLEDGMENTS This paper is dedicated to the memory of Jos´e Tormo, colleague and friend. We appreciate the generosity of R. Strong and P. Sun for providing manuscripts prior to publication. We thank the members of our laboratories for their help. Visit the Annual Reviews home page at www.annualreviews.org

LITERATURE CITED 1. Williams AF, Barclay AN. 1988. The immunoglobulin superfamily–domains for cell surface recognition. Annu. Rev. Immunol. 6:381–405 2. Drickamer K, Taylor ME. 1993. Biology of animal lectins. Annu. Rev. Cell Biol. 9:237–64 3. Weis WI, Taylor ME, Drickamer K. 1998. The C-type lectin superfamily in the immune system. Immunol. Rev. 163:19–34 4. Medzhitov R, Preston-Hurlburt P, Janeway CA Jr. 1997. A human homologue of the Drosophila Toll protein signals activation of adaptive immunity. Nature 388:394–97 5. Ljunggren HG, Karre K. 1990. In search of the ‘missing self’: MHC molecules and NK cell recognition. Immunol. Today 11:237–44 6. Karre K, Ljunggren HG, Piontek G, Kiessling R. 1986. Selective rejection of H-2deficient lymphoma variants suggests alternative immune defence strategy. Nature 319:675–78 7. Ljunggren HG, Karre K. 1985. Host resistance directed selectively against H-2-

8.

9.

10.

11.

12.

deficient lymphoma variants. Analysis of the mechanism. J. Exp. Med. 162:1745– 59 Shimizu Y, DeMars R. 1989. Demonstration by class I gene transfer that reduced susceptibility of human cells to natural killer cell-mediated lysis is inversely correlated with HLA class I antigen expression. Eur. J. Immunol. 19:447–51 Storkus WJ, Alexander J, Payne JA, Dawson JR, Cresswell P. 1989. Reversal of natural killing susceptibility in target cells expressing transfected class I HLA genes. Proc. Natl. Acad. Sci. USA 86: 2361–64 Colonna M, Borsellino G, Falco M, Ferrara GB, Strominger JL. 1993. HLA-C is the inhibitory ligand that determines dominant resistance to lysis by NK1- and NK2-specific natural killer cells. Proc. Natl. Acad. Sci. USA 90:12,000–4 Karlhofer FM, Ribaudo RK, Yokoyama WM. 1992. MHC class I alloantigen specificity of Ly-49+ IL-2-activated natural killer cells. Nature 358:66–70 Held W, Cado D, Raulet DH. 1996.

4 Feb 2002

12:4

876

Annu. Rev. Immunol. 2002.20:853-885. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

13.

14.

15. 16.

17.

18.

19.

20.

21.

AR

AR152-26.tex

AR152-26.SGM

LaTeX2e(2001/05/10)

P1: IKH

NATARAJAN ET AL. Transgenic expression of the Ly49A natural killer cell receptor confers class I major histocompatibility complex (MHC)-specific inhibition and prevents bone marrow allograft rejection. J. Exp. Med. 184:2037–41 Colonna M, Samaridis J. 1995. Cloning of immunoglobulin-superfamily members associated with HLA-C and HLA-B recognition by human natural killer cells. Science 268:405–8 Yokoyama WM, Jacobs LB, Kanagawa O, Shevach EM, Cohen DI. 1989. A murine T lymphocyte antigen belongs to a supergene family of type II integral membrane proteins. J. Immunol. 143:1379–86 Lanier LL. 1998. NK cell receptors. Annu. Rev. Immunol. 16:359–93 Moretta A, Bottino C, Vitale M, Pende D, Cantoni C, Mingari MC Biassoni R, Moretta L. 2001. Activating receptors and coreceptors involved in human natural killer cell-mediated cytolysis. Annu. Rev. Immunol. 19:197–223 Cudkowicz G, Bennett M. 1971. Peculiar immunobiology of bone marrow allografts. II. Rejection of parental grafts by resistant F1 hybrid mice. J. Exp. Med. 134:1513–28 Cudkowicz G, Bennett M. 1971. Peculiar immunobiology of bone marrow allografts. I. Graft rejection by irradiated responder mice. J. Exp. Med. 134:83–102 Murphy WJ, Kumar V, Bennett M. 1987. Rejection of bone marrow allografts by mice with severe combined immune deficiency (SCID). Evidence that natural killer cells can mediate the specificity of marrow graft rejection. J. Exp. Med. 165: 1212–17 Yu YY, George T, Dorfman JR, Roland J, Kumar V, Bennett M. 1996. The role of Ly49A and 5E6(Ly49C) molecules in hybrid resistance mediated by murine natural killer cells against normal T cell blasts. Immunity 4:67–76 Lanier LL. 2001. On guard–activating NK cell receptors. Nat. Immunol. 2:23–27

22. Lanier LL, Ruitenberg JJ, Phillips JH. 1988. Functional and biochemical analysis of CD16 antigen on natural killer cells and granulocytes. J. Immunol. 141:3478– 85 23. Moretta A, Poggi A, Pende D, Tripodi G, Orengo AM, Pella N, Augugliaro R, Bottino C, Ciccone E, Moretta L. 1991. CD69-mediated pathway of lymphocyte activation: anti-CD69 monoclonal antibodies trigger the cytolytic activity of different lymphoid effector cells with the exception of cytolytic T lymphocytes expressing T cell receptor alpha/beta. J. Exp. Med. 174:1393–98 24. Galandrini R, De Maria R, Piccoli M, Frati L, Santoni A. 1994. CD44 triggering enhances human NK cell cytotoxic functions. J. Immunol. 153:4399–4407 25. Yokoyama WM, Ryan JC, Hunter JJ, Smith HR, Stark M, Seaman WE. 1991. cDNA cloning of mouse NKR-P1 and genetic linkage with LY-49. Identification of a natural killer cell gene complex on mouse chromosome 6. J. Immunol. 147:3229–36 26. Siliciano RF, Pratt JC, Schmidt RE, Ritz J, Reinherz EL. 1985. Activation of cytolytic T lymphocyte and natural killer cell function through the T11 sheep erythrocyte binding protein. Nature 317:428–30. 27. Garni-Wagner BA, Purohit A, Mathew PA, Bennett M, Kumar V. 1993. A novel function-associated molecule related to non-MHC-restricted cytotoxicity mediated by activated natural killer cells and T cells. J. Immunol. 151:60–70 28. Tangye SG, Lazetic S, Woollatt E, Sutherland GR, Lanier LL, Phillips JH. 1999. Cutting edge: human 2B4, an activating NK cell receptor, recruits the protein tyrosine phosphatase SHP-2 and the adaptor signaling protein SAP. J. Immunol. 162:6981–85 29. Shibuya A, Campbell D, Hannum C, Yssel H, Franz-Bacon K, McClanahan T, Kitamura T, Nicholl J, Sutherland GR, Lanier LL, Phillips JH. 1996. DNAM-1,

4 Feb 2002

12:4

AR

AR152-26.tex

AR152-26.SGM

LaTeX2e(2001/05/10)

P1: IKH

NK CELL RECEPTORS

Annu. Rev. Immunol. 2002.20:853-885. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

30.

31.

32.

33.

34.

35.

a novel adhesion molecule involved in the cytolytic function of T lymphocytes. Immunity 4:573–81 Pende D, Parolini S, Pessino A, Sivori S, Augugliaro R, Morelli L, Marcenaro E, Accame L, Malaspina A, Biassoni R, Bottino C, Moretta L, Moretta A. 1999. Identification and molecular characterization of NKp30, a novel triggering receptor involved in natural cytotoxicity mediated by human natural killer cells. J. Exp. Med. 190:1505–16 Cantoni C, Bottino C, Vitale M, Pessino A, Augugliaro R, Malaspina A, Parolini S, Moretta L, Moretta A, Biassoni R. 1999. NKp44, a triggering receptor involved in tumor cell lysis by activated human natural killer cells, is a novel member of the immunoglobulin superfamily. J. Exp. Med. 189:787–96 Vitale M, Bottino C, Sivori S, Sanseverino L, Castriconi R, Marcenaro E, Augugliaro R, Moretta L, Moretta A. 1998. NKp44, a novel triggering surface molecule specifically expressed by activated natural killer cells, is involved in non-major histocompatibility complexrestricted tumor cell lysis. J. Exp. Med. 187:2065–72 Sivori S, Vitale M, Morelli L,Sanseverino L, Augugliaro R, Bottino C, Moretta L, Moretta A. 1997. p46, a novel natural killer cell-specific surface molecule that mediates cell activation. J. Exp. Med. 186: 1129–36 Falco M, Biassoni R, Bottino C, Vitale M, Sivori S, Augugliaro R, Moretta L, Moretta A. 1999. Identification and molecular cloning of p75/AIRM1, a novel member of the sialoadhesin family that functions as an inhibitory receptor in human natural killer cells. J. Exp. Med. 190: 793–802 Perussia B, Starr S, Abraham S, Fanning V, Trinchieri G. 1983. Human natural killer cells analyzed by B73.1, a monoclonal antibody blocking Fc receptor functions. I. Characterization of the lym-

36.

37.

38.

38a.

39.

40.

41.

42.

43.

877

phocyte subset reactive with B73.1. J. Immunol. 130:2133–41 Dustin ML, Sanders ME, Shaw S, Springer TA. 1987. Purified lymphocyte function-associated antigen 3 binds to CD2 and mediates T lymphocyte adhesion. J. Exp. Med. 165:677–92 van der Merwe PA, Brown MH, Davis SJ, Barclay AN. 1993. Affinity and kinetic analysis of the interaction of the cell adhesion molecules rat CD2 and CD48. Embo J. 12:4945–54 Brown MH, Boles K, van der Merwe PA, Kumar V, Mathew PA, Barclay AN. 1998. 2B4, the natural killer and T cell immunoglobulin superfamily surface protein, is a ligand for CD48. J. Exp. Med. 188:2083–90 Mandelboim O, Lieberman N, Lev M, Paul L, Arnon TI, Bushkin Y, Davis DM, Strominger JL, Yewdell JW, Porgador A. 2001. Recognition of haemagglutinins on virus-infected cells by NKp46 activates lysis by human NK cells. Nature 409:1055–60 Hanke T, Raulet DH. 2001. Cumulative inhibition of NK cells and T cells resulting from engagement of multiple inhibitory Ly49 receptors. J. Immunol. 166:3002–7 Ciccone E, Grossi CE, Velardi A. 1996. Opposing functions of activatory T-cell receptors and inhibitory NK-cell receptors on cytotoxic T cells. Immunol. Today 17:450–53 Lanier LL, Phillips JH. 1996. Inhibitory MHC class I receptors on NK cells and T cells. Immunol. Today 17:86–91 Mingari MC, Vitale C, Schiavetti F, Cambiaggi A, Bertone S, Zunino A, Ponte M. 1996. HLA-class I-specific inhibitory receptors of NK type on a subset of human T cells. Chem. Immunol. 64:135–45 Jiang W, Swiggard WJ, Heufler C, Peng M, Mirza A, Steinman RM, Nussenzweig MC. 1995. The receptor DEC-205 expressed by dendritic cells and thymic epithelial cells is involved in antigen processing. Nature 375:151–55

4 Feb 2002

12:4

Annu. Rev. Immunol. 2002.20:853-885. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

878

AR

AR152-26.tex

AR152-26.SGM

LaTeX2e(2001/05/10)

P1: IKH

NATARAJAN ET AL.

44. Geijtenbeek TB, van Vliet SJ, van Duijnhoven GC, Figdor CG, van Kooyk Y. 2001. DC-SIGN, a dentritic cell-specific HIV-1 receptor present in placenta that infects T cells in trans-a review. Placenta 22 (Suppl. A):S19–23 45. Hermanz-Falcon P, Arce I, Roda-Navarro P, Fernandez-Ruiz E. 2001. Cloning of human DECTIN-1, a novel C-type lectinlike receptor gene expressed on dendritic cells. Immunogenetics 53:288–95 46. Ariizumi K, Shen GL, Shikano S, Xu S, Ritter R III, Kumamoto T, Edelbaum D, Morita A, Bergstresser PR, Takashima A. 2000. Identification of a novel, dendritic cell-associated molecule, dectin-1, by subtractive cDNA cloning. J. Biol. Chem. 275:20157–67 47. Fan QR, Mosyak L, Winter CC, Wagtmann N, Long EO, Wiley DC. 1997. Structure of the inhibitory receptor for human natural killer cells resembles haematopoietic receptors. Nature 389:96–100 48. Snyder GA, Brooks AG, Sun PD. 1999. Crystal structure of the HLA-Cw3 allotype-specific killer cell inhibitory receptor KIR2DL2. Proc. Natl. Acad. Sci. USA 96:3864–69 49. Maenaka K, Juji T, Stuart DI, Jones EY. 1999. Crystal structure of the human p58 killer cell inhibitory receptor (KIR2DL3) specific for HLA-Cw3-related MHC class I. Structure Fold Des. 7:391–98 50. Boyington JC, Motyka SA, Schuck P, Brooks AG, Sun PD. 2000. Crystal structure of an NK cell immunoglobulin-like receptor in complex with its class I MHC ligand. Nature 405:537–43 51. Fan QR, Long EO, Wiley DC. 2001. Crystal structure of the human natural killer cell inhibitory receptor KIR2DL1HLA-Cw4 complex. Nat. Immunol. 2: 452–60 52. Chapman TL, Heikema AP, West AP Jr, Bjorkman PJ. 2000. Crystal structure and ligand binding properties of the D1D2 region of the inhibitory receptor LIR-1 (ILT2). Immunity 13:727–36

53. Boyington JC, Riaz AN, Patamawenu A, Coligan JE, Brooks AG, Sun PD. 1999. Structure of CD94 reveals a novel C-type lectin fold: implications for the NK cellassociated CD94/NKG2 receptors. Immunity 10:75–82 54. Natarajan K, Sawicki MW, Margulies DH, Mariuzza RA. 2000. Crystal structure of human CD69: a C-type lectin-like activation marker of hematopoietic cells. Biochemistry 39:14779–86 55. Llera AS, Viedma F, Sanchez-Madrid F, Tormo J. 2001. Crystal structure of the Ctype lectin-like domain from the human hematopoietic cell receptor CD69. J. Biol. Chem. 276:7312–19 56. Wolan DW, Teyton L, Rudolph MG, Villmow B, Bauer S, Busch DH, Wilson IA. 2001. Crystal structure of the murine NK cell-activating receptor NKG2D at 1.95 A. Nat. Immunol. 2:248–54 57. Tormo J, Natarajan K, Margulies DH, Mariuzza RA. 1999. Crystal structure of a lectin-like natural killer cell receptor bound to its MHC class I ligand. Nature 402:623–31 58. Li P, Morris DL, Willcox BE, Steinle A, Spies T, Strong RK. 2001. Complex structure of the activating immunoreceptor NKG2D and its MHC class I-like ligand MICA. Nat. Immunol. 2:443–51 59. Wagtmann N, Biassoni R, Cantoni C, Verdiani S, Malnati MS, Vitale M, Bottino C, Moretta L, Moretta A, Long EO. 1995. Molecular clones of the p58 NK cell receptor reveal immunoglobulin-related molecules with diversity in both the extra- and intracellular domains. Immunity 2:439–49 60. Burshtyn DN, Scharenberg AM, Wagtmann N, Rajagopalan S, Berrada K, Yi T, Kinet JP, Long EO. 1996. Recruitment of tyrosine phosphatase HCP by the killer cell inhibitor receptor. Immunity 4:77–85 61. Campbell KS, Dessing M, Lopez-Botet M, Cella M, Colonna M. 1996. Tyrosine phosphorylation of a human killer inhibitory receptor recruits protein tyrosine

4 Feb 2002

12:4

AR

AR152-26.tex

AR152-26.SGM

LaTeX2e(2001/05/10)

P1: IKH

NK CELL RECEPTORS

62.

Annu. Rev. Immunol. 2002.20:853-885. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

63.

64.

65.

66.

67.

68.

69.

70.

phosphatase 1C. J. Exp. Med. 184:93– 100 Long EO. 1999. Regulation of immune responses through inhibitory receptors. Annu. Rev. Immunol. 17:875–904 Wu J, Cherwinski H, Spies T, Phillips JH, Lanier LL. 2000. DAP10 and DAP12 form distinct, but functionally cooperative, receptor complexes in natural killer cells. J. Exp. Med. 192:1059–68 Tomasello E, Olcese L, Vely F, Geourgeon C, Blery M, Moqrich A, Gautheret D, Djabali M, Mattei MG, Vivier E. 1998. Gene structure, expression pattern, and biological activity of mouse killer cell activating receptor-associated protein (KARAP)/DAP-12. J. Biol. Chem. 273:34,115–19 Wu J, Song Y, Bakker AB, Bauer S, Spies T, Lanier LL, Phillips JH. 1999. An activating immunoreceptor complex formed by NKG2D and DAP10. Science 285:730–32 Chang C, Dietrich J, Harpur AG, Lindquist JA, Haude A, Loke YW, King A, Colonna M, Trowsdale J, Wilson MJ. 1999. Cutting edge: KAP10, a novel transmembrane adapter protein genetically linked to DAP12 but with unique signaling properties. J. Immunol. 163:4651–54 Ferrini S, Cambiaggi A, Meazza R, Sforzini S, Marciano S, Mingari MC, Moretta L. 1994. T cell clones expressing the natural killer cell-related p58 receptor molecule display heterogeneity in phenotypic properties and p58 function. Eur. J. Immunol. 24:2294–98 Battistini L, Borsellino G, Sawicki G, Poccia F, Salvetti M, Ristori G, Brosnan CF. 1997. Phenotypic and cytokine analysis of human peripheral blood gamma delta T cells expressing NK cell receptors. J. Immunol. 159:3723–30 D0 Andrea A, Lanier LL. 1998. Killer cell inhibitory receptor expression by T cells. Curr. Top. Microbiol. Immunol. 230:25– 39 Litwin V, Gumperz J, Parham P, Phillips

71.

72.

73.

74.

75.

76.

77.

78.

879

JH, Lanier LL. 1993. Specificity of HLA class I antigen recognition by human NK clones: evidence for clonal heterogeneity, protection by self and non-self alleles, and influence of the target cell type. J. Exp. Med. 178:1321–36 Wagtmann N, Rajagopalan S, Winter CC, Peruzzi M, Long EO. 1995. Killer cell inhibitory receptors specific for HLA-C and HLA-B identified by direct binding and by functional transfer. Immunity 3:801–9 Dohring C, Colonna M. 1996. Human natural killer cell inhibitory receptors bind to HLA class I molecules. Eur. J. Immunol. 26:365–69 Fan QR, Garboczi DN, Winter CC, Wagtmann N, Long EO, Wiley DC. 1996. Direct binding of a soluble natural killer cell inhibitory receptor to a soluble human leukocyte antigen-Cw4 class I major histocompatibility complex molecule. Proc. Natl. Acad. Sci. USA 93:7178–83 Mandelboim O, Reyburn HT, Sheu EG, Vales-Gomez M, Davis DM, Pazmany L, Strominger JL. 1997. The binding site of NK receptors on HLA-C molecules. Immunity 6:341–50 Vales-Gomez M, Reyburn HT, Mandelboim M, Strominger JL. 1998. Kinetics of interaction of HLA-C ligands with natural killer cell inhibitory receptors. Immunity 9:337–44 Vales-Gomez M, Reyburn H, Strominger J. 2000. Interaction between the human NK receptors and their ligands. Crit. Rev. Immunol. 20:223–44 Cella M, Longo A, Ferrara GB, Strominger JL, Colonna M. 1994. NK3-specific natural killer cells are selectively inhibited by Bw4-positive HLA alleles with isoleucine 80. J. Exp. Med. 180:1235– 42 Gumperz JE, Litwin V, Phillips JH, Lanier LL, Parham P. 1995. The Bw4 public epitope of HLA-B molecules confers reactivity with natural killer cell clones that express NKB1, a putative HLA receptor. J. Exp. Med. 181:1133–44

4 Feb 2002

12:4

Annu. Rev. Immunol. 2002.20:853-885. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

880

AR

AR152-26.tex

AR152-26.SGM

LaTeX2e(2001/05/10)

P1: IKH

NATARAJAN ET AL.

79. Luque I, Solana R, Galiani MD, Gonzalez R, Garcia F, Lopez de Castro JA, Pena J. 1996. Threonine 80 on HLA-B27 confers protection against lysis by a group of natural killer clones. Eur. J. Immunol. 26:1974–77 80. Rojo S, Wagtmann N, Long EO. 1997. Binding of a soluble p70 killer cell inhibitory receptor to HLA-B∗ 5101: requirement for all three p70 immunoglobulin domains. Eur. J. Immunol. 27:568–71 81. Dohring C, Scheidegger D, Samaridis J, Cella M, Colonna M. 1996. A human killer inhibitory receptor specific for HLA-A1,2. J. Immunol. 156:3098–3101 82. Pende D, Biassoni R, Cantoni C, Verdiani S, Falco M, di Donato C, Accame L, Bottino C, Moretta A, Moretta L. 1996. The natural killer cell receptor specific for HLA-A allotypes: a novel member of the p58/p70 family of inhibitory receptors that is characterized by three immunoglobulin-like domains and is expressed as a 140-kD disulphidelinked dimer. J. Exp. Med. 184:505–18 83. Biassoni R, Pessino A, Malaspina A, Cantoni C, Bottino C, Sivori S, Moretta L, Moretta A. 1997. Role of amino acid position 70 in the binding affinity of p50.1 and p58.1 receptors for HLA-Cw4 molecules. Eur. J. Immunol. 27:3095–99 84. Malnati MS, Peruzzi M, Parker KC, Biddison WE, Ciccone E, Moretta A, Long EO. 1995. Peptide specificity in the recognition of MHC class I by natural killer cell clones. Science 267:1016–18 85. Mandelboim O, Reyburn H, ValesGomez M, Pazmany L, Colonna M, Borsellino G, Strominger J. 1996. Protection from lysis by natural killer cells of group 1 and 2 specificity is mediated by residue 80 in human histocompatibility leukocyte antigen C alleles and also occurs with empty major histocompatibility complex molecules. J. Exp. Med. 184:913–22 86. Bork P, Holm L, Sander C. 1994. The immunoglobulin fold. Structural classifica-

87.

88.

89.

90.

91.

92.

93.

94.

95.

tion, sequence patterns and common core. J. Mol. Biol. 242:309–20 de Vos AM, Ultsch M, Kossiakoff AA. 1992. Human growth hormone and extracellular domain of its receptor: crystal structure of the complex. Science 255: 306–12 Livnah O, Stura EA, Middleton SA, Johnson DL, Jolliffe LK, Wilson IA. 1999. Crystallographic evidence for preformed dimers of erythropoietin receptor before ligand activation. Science 283: 987–90 Somers W, Ultsch M, De Vos AM, Kossiakoff AA. 1994. The X-ray structure of a growth hormone-prolactin receptor complex. Nature 372:478–81 Jones EY, Davis SJ, Williams AF, Harlos K, Stuart DI. 1992. Crystal structure ˚ resolution of a soluble form of at 2.8 A the cell adhesion molecule CD2. Nature 360:232–39 Ryu SE, Truneh A, Sweet RW, Hendrickson WA. 1994. Structures of an HIV and MHC binding fragment from human CD4 as refined in two crystal lattices. Structure 2:59–74 Wang JH, Yan YW, Garrett TP, Liu JH, Rodgers DW, Garlick RL, Tarr GE, Husain Y, Reinherz EL, Harrison SC. 1990. Atomic structure of a fragment of human CD4 containing two immunoglobulinlike domains. Nature 348:411–18 Winter CC, Long EO. 1997. A single amino acid in the p58 killer cell inhibitory receptor controls the ability of natural killer cells to discriminate between the two groups of HLA-C allotypes. J. Immunol. 158:4026–28 Winter CC, Gumperz JE, Parham P, Long EO, Wagtmann N. 1998. Direct binding and functional transfer of NK cell inhibitory receptors reveal novel patterns of HLA-C allotype recognition. J. Immunol. 161:571–77 Garboczi DN, Ghosh P, Utz U, Fan QR, Biddison WE, Wiley DC. 1996. Structure of the complex between human T-cell

4 Feb 2002

12:4

AR

AR152-26.tex

AR152-26.SGM

LaTeX2e(2001/05/10)

P1: IKH

NK CELL RECEPTORS

96.

Annu. Rev. Immunol. 2002.20:853-885. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

97.

98.

99.

100.

101.

102.

103.

104.

receptor, viral peptide and HLA-A2. Nature 384:134–41 Ding YH, Smith KJ, Garboczi DN, Utz U, Biddison WE, Wiley DC. 1998. Two human T cell receptors bind in a similar diagonal mode to the HLA-A2/Tax peptide complex using different TCR amino acids. Immunity 8:403–11 Garcia KC, Degano M, Stanfield RL, Brunmark A, Jackson MR, Peterson PA, Teyton L, Wilson IA. 1996. An alpha˚ and beta T cell receptor structure at 2.5 A its orientation in the TCR-MHC complex. Science 274:209–19 Wilson IA, Jolliffe LK. 1999. The structure, organization, activation and plasticity of the erythropoietin receptor. Curr. Opin. Struct. Biol. 9:696–704 Garcia KC, Teyton L, Wilson IA. 1999. Structural basis of T cell recognition. Annu. Rev. Immunol. 17:369–97 Reinherz EL, Tan K, Tang L, Kern P, Liu J, Xiong Y, Hussey RE, Smolyar A, Hare B, Zhang R, Joachimiak A, Chang HC, Wagner G, Wang J. 1999. The crystal structure of a T cell receptor in complex with peptide and MHC class II. Science 286:1913–21 Hennecke J, Carfi A, Wiley DC. 2000. Structure of a covalently stabilized complex of a human alphabeta T- cell receptor, influenza HA peptide and MHC class II molecule, HLA-DR1. Embo J. 19:5611– 24 Lawrence MC, Colman PM. 1993. Shape complementarity at protein/protein interfaces. J. Mol. Biol. 234:946–50 Wang JH, Smolyar A, Tan K, Liu JH, Kim M, Sun ZY, Wagner G, Reinherz EL. 1999. Structure of a heterophilic adhesion complex between the human CD2 and CD58 (LFA-3) counterreceptors. Cell 97:791–803 Degano M, Garcia KC, Apostolopoulos V, Rudolph MG, Teyton L, Wilson IA. 2000. A functional hot spot for antigen recognition in a superagonist TCR/MHC complex. Immunity 12:251–61

881

105. Ding YH, Baker BM, Garboczi DN, Biddison WE, Wiley DC. 1999. Four A6-TCR/peptide/HLA-A2 structures that generate very different T cell signals are nearly identical. Immunity 11:45–56 106. Rajagopalan S, Long EO. 1997. The direct binding of a p58 killer cell inhibitory receptor to human histocompatibility leukocyte antigen (HLA)-Cw4 exhibits peptide selectivity. J. Exp. Med. 185:1523–28 107. Peruzzi M, Parker KC, Long EO, Malnati MS. 1996. Peptide sequence requirements for the recognition of HLA-B∗ 2705 by specific natural killer cells. J. Immunol. 157:3350–56 108. Fan QR, Wiley DC. 1999. Structure of human histocompatibility leukocyte antigen (HLA)-Cw4, a ligand for the KIR2D natural killer cell inhibitory receptor. J. Exp. Med. 190:113–23 109. Fanger NA, Borges L, Cosman D. 1999. The leukocyte immunoglobulin-like receptors (LIRs): a new family of immune regulators. J. Leukoc. Biol. 66:231–36 110. Borges L, Hsu ML, Fanger N, Kubin M, Cosman D. 1997. A family of human lymphoid and myeloid Ig-like receptors, some of which bind to MHC class I molecules. J. Immunol. 159:5192–96 111. Colonna M, Navarro F, Bellon T, Llano M, Garcia P, Samaridis J, Angman L, Cella M, Lopez-Botet M. 1997. A common inhibitory receptor for major histocompatibility complex class I molecules on human lymphoid and myelomonocytic cells. J. Exp. Med. 186:1809–18 112. Cella M, Dohring C, Samaridis J, Dessing M, Brockhaus M, Lanzavecchia A, Colonna M. 1997. A novel inhibitory receptor (ILT3) expressed on monocytes, macrophages, and dendritic cells involved in antigen processing. J. Exp. Med. 185: 1743–51 113. Samaridis J, Colonna M. 1997. Cloning of novel immunoglobulin superfamily receptors expressed on human myeloid and lymphoid cells: structural evidence for

4 Feb 2002

12:4

882

114.

Annu. Rev. Immunol. 2002.20:853-885. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

115.

116.

117.

118.

119.

120.

121.

AR

AR152-26.tex

AR152-26.SGM

LaTeX2e(2001/05/10)

P1: IKH

NATARAJAN ET AL. new stimulatory and inhibitory pathways. Eur. J. Immunol. 27:660–65 Cella M, Nakajima H, Facchetti F, Hoffmann T, Colonna M. 2000. ILT receptors at the interface between lymphoid and myeloid cells. Curr. Top. Microbiol. Immunol. 251:161–66 Dietrich J, Nakajima H, Colonna M. 2000. Human inhibitory and activating Ig-like receptors which modulate the function of myeloid cells. Microbes Infect. 2:323–29 Fanger NA, Cosman D, Peterson L, Braddy SC, Maliszewski CR, Borges L. 1998. The MHC class I binding proteins LIR-1 and LIR-2 inhibit Fc receptor–mediated signaling in monocytes. Eur. J. Immunol. 28:3423–34 Mizuno H, Fujimoto Z, Koizumi M, Kano H, Atoda H, Morita T. 1999. Crystal structure of coagulation factor IX-binding protein from habu snake venom at 2.6 ˚ implication of central loop swapping A: based on deletion in the linker region. J. Mol. Biol. 289:103–12 Mizuno H, Fujimoto Z, Koizumi M, Kano H, Atoda H, Morita T. 1997. Structure of coagulation factors IX/Xbinding protein, a heterodimer of C-type lectin domains. Nat. Struct. Biol. 4:438– 41 Shi J, Ghirlando R, Beavil RL, Beavil AJ, Keown MB, Young RJ, Owens RJ, Sutton BJ, Gould HJ. 1997. Interaction of the low-affinity receptor CD23/Fc epsilonRII lectin domain with the Fc epsilon3-4 fragment of human immunoglobulin E. Biochemistry 36:2112–22 Loewen MC, Gronwald W, Sonnichsen FD, Sykes BD, Davies PL. 1998. The icebinding site of sea raven antifreeze protein is distinct from the carbohydrate-binding site of the homologous C-type lectin. Biochemistry 37:17745–53 Gronwald W, Loewen MC, Lix B, Daugulis AJ, Sonnichsen FD, Davies PL, Sykes BD. 1998. The solution structure of type II antifreeze protein reveals a new

122.

123.

124.

125.

126.

127.

128.

129.

130.

131.

132.

member of the lectin family. Biochemistry 37:4712–21 Bertrand JA, Pignol D, Bernard JP, Verdier JM, Dagorn JC, Fontecilla-Camps JC. 1996. Crystal structure of human lithostathine, the pancreatic inhibitor of stone formation. Embo J. 15:2678–84 Hakansson K, Reid KB. 2000. Collectin structure: a review. Protein Sci. 9:1607– 17 Taylor ME. 2001. Structure and function of the macrophage mannose receptor. Results Probl. Cell Differ. 33:105–21 Linehan SA, Martinez-Pomares L, Gordon S. 2000. Mannose receptor and scavenger receptor: two macrophage pattern recognition receptors with diverse functions in tissue homeostasis and host defense. Adv. Exp. Med. Biol. 479:1–14 Yokoyama WM. 1999. Natural killer cells. In Fundamental Immunology, ed. WE Paul, pp. 575–603. Philadelphia: Lippincott-Raven McQueen KL, Lohwasser S, Takei F, Mager DL. 1999. Expression analysis of new Ly49 genes: most transcripts of Ly49j lack the transmembrane domain. Immunogenetics 49:685–91 Makrigiannis AP, Pau AT, Saleh A, Winkler-Pickett R, Ortaldo JR, Anderson SK. 2001. Class I MHC-binding characteristics of the 129/j ly49 repertoire. J. Immunol. 166:5034–43 Westgaard IH, Berg SF, Orstavik S, Fossum S, Dissen E. 1998. Identification of a human member of the Ly-49 multigene family. Eur. J. Immunol. 28:1839–46 Trowsdale J, Barten R, Haude A, Stewart CA, Beck S, Wilson MJ. 2001. The genomic context of natural killer receptor extended gene families. Immunol. Rev. 181:20–38 Mager DL, McQueen KL, Wee V, Freeman JD. 2001. Evolution of natural killer cell receptors: coexistence of functional Ly49 and KIR genes in baboons. Curr. Biol. 11:626–30 Orihuela M, Margulies DH, Yokoyama

4 Feb 2002

12:4

AR

AR152-26.tex

AR152-26.SGM

LaTeX2e(2001/05/10)

P1: IKH

NK CELL RECEPTORS

Annu. Rev. Immunol. 2002.20:853-885. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

133.

134.

135.

136.

137.

138.

139.

140.

WM. 1996. The natural killer cell receptor Ly-49A recognizes a peptide-induced conformational determinant on its major histocompatibility complex class I ligand. Proc. Natl. Acad. Sci. USA 93:11792–97 Karlhofer FM, Hunziker R, Reichlin A, Margulies DH, Yokoyama WM. 1994. Host MHC class I molecules modulate in vivo expression of a NK cell receptor. J. Immunol. 153:2407–16 Natarajan K, Boyd LF, Schuck P, Yokoyama WM, Eliat D, Margulies DH. 1999. Interaction of the NK cell inhibitory receptor Ly49A with H-2Dd: identification of a site distinct from the TCR site. Immunity 11:591–601 Fremont DH, Rees WA, Kozono H. 1996. Biophysical studies of T-cell receptors and their ligands. Curr. Opin. Immunol. 8:93–100 Margulies DH. 1997. Interactions of TCRs with MHC-peptide complexes: a quantitative basis for mechanistic models. Curr. Opin. Immunol. 9:390–95 Plaksin D, Polakova K, McPhie P, Margulies DH. 1997. A three-domain T cell receptor is biologically active and specifically stains cell surface MHC/peptide complexes. J. Immunol. 158:2218– 27 Chang C, Rodriguez A, Carretero M, Lopez-Botet M, Phillips JH, Lanier LL. 1995. Molecular characterization of human CD94: a type II membrane glycoprotein related to the C-type lectin superfamily. Eur. J. Immunol. 25:2433–37 Le Drean E, Vely F, Olcese L, Cambiaggi A, Guia S, Krystal G, Gervois N, Moretta A, Jotereau F, Vivier E. 1998. Inhibition of antigen-induced T cell response and antibody-induced NK cell cytotoxicity by NKG2A: association of NKG2A with SHP-1 and SHP-2 protein-tyrosine phosphatases. Eur. J. Immunol. 28:264–76 Olcese L, Cambiaggi A, Semenzato G, Bottino C, Moretta A, Vivier E. 1997. Human killer cell activatory receptors for MHC class I molecules are included in a

141.

142.

143.

144.

145.

146.

147.

148.

149.

883

multimeric complex expressed by natural killer cells. J. Immunol. 158:5083–86 Lanier LL, Corliss B, Wu J, Phillips JH. 1998. Association of DAP12 with activating CD94/NKG2C NK cell receptors. Immunity 8:693–701 Soderstrom K, Corliss B, Lanier LL, Phillips JH. 1997. CD94/NKG2 is the predominant inhibitory receptor involved in recognition of HLA-G by decidual and peripheral blood NK cells. J. Immunol. 159:1072–75 Brooks AG, Posch PE, Scorzelli CJ, Borrego F, Coligan JE. 1997. NKG2A complexed with CD94 defines a novel inhibitory natural killer cell receptor. J. Exp. Med. 185:795–800 Posch PE, Borrego F, Brooks AG, Coligan JE. 1998. HLA-E is the ligand for the natural killer cell CD94/NKG2 receptors. J. Biomed. Sci. 5:321–31 Vance RE, Kraft JR, Altman JD, Jensen PE, Raulet DH. 1998. Mouse CD94/ NKG2A is a natural killer cell receptor for the nonclassical major histocompatibility complex (MHC) class I molecule Qa-1(b). J. Exp. Med. 188:1841–48 Kurepa Z, Hasemann CA, Forman J. 1998. Qa-1b binds conserved class I leader peptides derived from several mammalian species. J. Exp. Med. 188:973–78 Braud V, Jones EY, McMichael A. 1997. The human major histocompatibility complex class Ib molecule HLA-E binds signal sequence-derived peptides with primary anchor residues at positions 2 and 9. Eur. J. Immunol. 27:1164–69 Borrego F, Ulbrecht M, Weiss EH, Coligan JE, Brooks AG. 1998. Recognition of human histocompatibility leukocyte antigen (HLA)-E complexed with HLA class I signal sequence-derived peptides by CD94/NKG2 confers protection from natural killer cell-mediated lysis. J. Exp. Med. 187:813–18 Braud VM, Allan DS, O’Callaghan CA, Soderstrom K, D’Andrea A, Ogg GS, Lazetic S, Young NT, Bell JI, Phillips JH,

4 Feb 2002

12:4

884

Annu. Rev. Immunol. 2002.20:853-885. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

150.

151.

152.

153.

154.

155.

156.

157.

AR

AR152-26.tex

AR152-26.SGM

LaTeX2e(2001/05/10)

P1: IKH

NATARAJAN ET AL. Lanier LL, McMichael AJ. 1998. HLA-E binds to natural killer cell receptors CD94/NKG2A, B and C. Nature 391: 795–99 Lee N, Llano M, Carretero M, Ishitani A, Navarro F, Lopez-Botet M, Geraghty DE. 1998. HLA-E is a major ligand for the natural killer inhibitory receptor CD94/NKG2A. Proc. Natl. Acad. Sci. USA 95:5199–5204 Brooks AG, Borrego F, Posch PE, Patamawenu A, Scorzelli CJ, Ulbrecht M, Weiss EH, Coligan JE. 1999. Specific recognition of HLA-E, but not classical, HLA class I molecules by soluble CD94/NKG2A and NK cells. J. Immunol. 162:305–13 Vales-Gomez M, Reyburn HT, Erskine RA, Lopez-Botet M, Strominger JL. 1999. Kinetics and peptide dependency of the binding of the inhibitory NK receptor CD94/NKG2-A and the activating receptor CD94/NKG2-C to HLA-E. Embo J. 18:4250–60 Bauer S, Groh V, Wu J, Steinle A, Phillips JH, Lanier LL, Spies T. 1999. Activation of NK cells and T cells by NKG2D, a receptor for stress-inducible MICA. Science 285:727–29 Groh V, Rhinehart R, Secrist H, Bauer S, Grabstein KH, Spies T. 1999. Broad tumor-associated expression and recognition by tumor-derived gamma delta T cells of MICA and MICB. Proc. Natl. Acad. Sci. USA 96:6879–84 Groh V, Steinle A, Bauer S, Spies T. 1998. Recognition of stress-induced MHC molecules by intestinal epithelial gammadelta T cells. Science 279:1737– 40 Groh V, Bahram S, Bauer S, Herman A, Beauchamp M, Spies T. 1996. Cell stress-regulated human major histocompatibility complex class I gene expressed in gastrointestinal epithelium. Proc. Natl. Acad. Sci. USA 93:12,445–50 Cosman D, Mullberg J, Sutherland CL, Chin W, Armitage R, Fanslow W, Kubin

158.

159.

160.

161.

162.

163.

164.

165.

166.

M, Chalupny NJ. 2001. ULBPs, novel MHC class I-related molecules, bind to CMV glycoprotein UL16 and stimulate NK cytotoxicity through the NKG2D receptor. Immunity 14:123–33 Cerwenka A, Bakker AB, McClanahan T, Wagner J, Wu J, Phillips JH, Lanier LL. 2000. Retinoic acid early inducible genes define a ligand family for the activating NKG2D receptor in mice. Immunity 12:721–27 Diefenbach A, Jamieson AM, Liu SD, Shastri N, Raulet DH. 2000. Ligands for the murine NKG2D receptor: expression by tumor cells and activation of NK cells and macrophages. Nat. Immunol. 1:119– 26 Diefenbach A, Jensen ER, Jamieson AM, Raulet DH. 2001. Rae1 and H60 ligands of the NKG2D receptor stimulate tumour immunity. Nature 413:165–71 Kastrup JS, Nielsen BB, Rasmussen H, Holtet TL, Graversen JH, Etzerodt M, Thogersen HC, Larsen IK. 1998. Structure of the C-type lectin carbohydrate recognition domain of human tetranectin. Acta Crystallogr. D Biol. Crystallogr. 54: 757–66 Nielsen BB, Kastrup JS, Rasmussen H, Holtet TL, Graversen JH, Etzerodt M, Thogersen HC, Larsen IK. 1997. Crystal structure of tetranectin, a trimeric plasminogen-binding protein with an alpha-helical coiled coil. FEBS Lett. 412:388–96 Lanier LL. 1998. Activating and inhibitory NK cell receptors. Adv. Exp. Med. Biol. 452:13–18 Poget SF, Legge GB, Proctor MR, Butler PJ, Bycroft M, Williams RL. 1999. The structure of a tunicate C-type lectin from Polyandrocarpa misakiensis complexed with D-galactose. J. Mol. Biol. 290:867–79 Conte LL, Chothia C, Janin J. 1999. The atomic structure of protein-protein recognition sites. J. Mol. Biol. 285:2177–98 Gao GF, Tormo J, Gerth UC, Wyer JR, McMichael AJ, Stuart DI, Bell JI,

4 Feb 2002

12:4

AR

AR152-26.tex

AR152-26.SGM

LaTeX2e(2001/05/10)

P1: IKH

NK CELL RECEPTORS

Annu. Rev. Immunol. 2002.20:853-885. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

167.

168.

169.

170.

171.

171a.

172.

Jones EY, Jakobsen BK. 1997. Crystal structure of the complex between human CD8alpha(alpha) and HLA-A2. Nature 387:630–34 Kern PS, Teng MK, Smolyar A, Liu JH, Liu J, Hussey RE, Spoerl R, Chang HC, Reinherz EL, Wang JH. 1998. Structural basis of CD8 coreceptor function revealed by crystallographic analysis of a murine CD8alphaalpha ectodomain fragment in complex with H-2Kb. Immunity 9:519–30 Parham P. 2000. NK cell receptors: of missing sugar and missing self. Curr. Biol. 10:R195–97 Chung DH, Dorfman J, Plaksin D, Natarajan K, Belyakov IM, Hunziker R, Berzofsky JA, Yokoyama WM, Mage MG, Margulies DH. 1999. NK and CTL recognition of a single chain H-2Dd molecule: distinct sites of H-2Dd interact with NK and TCR. J. Immunol. 163:3699–3708 Chung DH, Natarajan K, Boyd LF, Tormo J, Mariuzza RA, Yokoyama WM, Margulies DH. 2000. Mapping the ligand of the NK inhibitory receptor Ly49A on living cells. J. Immunol. 165:6922–32 Matsumoto N, Mitsuki M, Tajima K, Yokoyama WM, Yamamoto K. 2001. The functional binding site for the Ctype lectin-like natural killer cell receptor Ly49A spans three domains of its major histocompatibility complex class I ligand. J. Exp. Med. 193:147–58 Wang J, Whitman MC, Natarajan K, Tormo J, Marriuzza RA, Margulies DH. 2002. Binding of the NK cell inhibitory receptor Ly49A to its MHC-I ligand: crucial contacts include both H-2Dd and β2-m. J. Biol. Chem. In press Brennan J, Mahon G, Mager DL, Jefferies WA, Takei F. 1996. Recognition of class I major histocompatibility complex molecules by Ly-49: specificities and domain interactions. J. Exp. Med. 183:1553–59

885

173. Li P, Willie ST, Bauer S, Morris DL, Spies T, Strong RK. 1999. Crystal structure of the MHC class I homolog MICA, a gammadelta T cell ligand. Immunity 10:577–84 174. O’Callaghan CA, Tormo J, Willcox BE, Braud VM, Jakobsen BK, Stuart DI, McMichael AJ, Bell JI, Jones EY. 1998. Structural features impose tight peptide binding specificity in the nonclassical MHC molecule HLA-E. Mol. Cell. 1:531–41 175. Kraft JR, Vance RE, Pohl J, Martin AM, Raulet DH, Jensen PE. 2000. Analysis of Qa-1(b) peptide binding specificity and the capacity of CD94/NKG2A to discriminate between Qa-1-peptide complexes. J. Exp. Med. 192:613–24 176. Kubagawa H, Cooper MD, Chen CC, Ho LH, Alley TL, Hurez V, Tun T, Uehara T, Shimada T, Burrows PD. 1999. Paired immunoglobulin-like receptors of activating and inhibitory types. Curr. Top. Microbiol. Immunol. 244:137– 49 177. Kubagawa H, Chen CC, Ho LH, Shimada TS, Gartland L, Mashburn C, Uehara T, Ravetch JV, Cooper MD. 1999. Biochemical nature and cellular distribution of the paired immunoglobulinlike receptors, PIR-A and PIR-B. J. Exp. Med. 189:309–18 178. Andre P, Biassoni R, Colonna M, Cosman D, Lanier LL, Long EO, LopezBotet M, Moretta A, Moretta L, Parham P, Trowsdale J, Vivier E, Wagtmann N, Wilson MJ. 2001. New nomenclature for MHC receptors. Nat. Immunol. 2:661 179. Weis WI, Kahn R, Fourme R, Drickamer K, Hendrickson WA. 1991. Structure of the calcium-dependent lectin domain from a rat mannose-binding protein determined by MAD phasing. Science 254:1608–15 180. Gouet P, Courcelle E, Stuart DI, Metoz F. 1999. ESPript: analysis of multiple sequence alignments in PostScript. Bioinformatics 15:305–8

Annu. Rev. Immunol. 2002.20:853-885. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

7 Feb 2002

9:12

AR

AR152-26-COLOR.tex

AR152-26-COLOR.SGM

LaTeX2e(2002/01/18)

P1: GDL

Figure 1 Three-dimensional structures of KIR2DL and LIR-1. (A) Ribbon diagram of KIR2DL1 (PDB entry code 1NKR) (47). (B) Ribbon diagram of LIR-1 (1GOX) (52). β-strands are shown in green or blue and α-helices in red. The secondary structural elements are labeled. The region of KIR2DL1 in the interface with HLA-Cw4 in the KIR2DL1/HLA-Cw4 complex structure (51) is indicated by a pink transparent ellipse. This region comprises loops A0 B, CC0 and EF of the D1 domain, the hinge loop, and loops BC and FG of the D2 domain. The predicted interface region of LIR-1 with the cytomegalovirus-encoded MHC-I homolog UL18 is also represented by a pink transparent ellipse.

9:12

AR

AR152-26-COLOR.tex

AR152-26-COLOR.SGM

LaTeX2e(2002/01/18)

P1: GDL

Annu. Rev. Immunol. 2002.20:853-885. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

7 Feb 2002

Figure 2 Structure-based amino acid sequence alignment of KIRs (A) and C-type lectin-like molecules (B). (A) Amino acid sequences for the regions of KIR2DL1 (47), KIR2DL2 (48), and KIR2DL3 (49) for which X-ray structure has been determined are shown with secondary structure and strand names indicated. (B) Similar alignment with display of the secondary structure of Ly49A for Ly49A (57), Ly49I (N Dimasi, MW Sawicki, LN Reineck, Y Li, K Natarajan, DH Margulies & RA Mariuzza, unpublished data), hCD94 (53), hNKG2D (58), mNKG2D (56), hCD69 (54, 55), and MBP-A (179) is shown. Figure was generated using Espript (180).

Annu. Rev. Immunol. 2002.20:853-885. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

7 Feb 2002

9:12

AR

AR152-26-COLOR.tex

AR152-26-COLOR.SGM

LaTeX2e(2002/01/18)

P1: GDL

Figure 3 Structures of KIR2DL/HLA-C complexes and the basis for allelic specificity. (A) Ribbon diagram of the KIR2DL2/HLA-Cw3 complex (PDB entry code 1EFX) (50). (B) Ribbon diagram of the KIR2DL1/HLA-Cw4 complex (1IM9) (51). The bound KIR molecules are colored in magenta. The α1, α2 and α3 domains of the HLA-Cw3 and HLA-Cw4 heavy chains are yellow, the peptides are blue, and β2m molecules are gray. (C) Interaction of Asn80 of HLA-Cw3 (yellow) with Lys44 of KIR2DL2 (magenta). (D) Specificity pocket of KIR2DL1 (magenta) that accommodates Lys80 of HLA-Cw4 (yellow). Hydrogen bonds are illustrated as black dots. (An animation of the KIR2DL2/HLA-Cw3 complex structure is offered as supplementary material in the online version of this chapter or at http://www.annualreviews.org/.)

9:12

AR

AR152-26-COLOR.tex

AR152-26-COLOR.SGM

LaTeX2e(2002/01/18)

P1: GDL

Annu. Rev. Immunol. 2002.20:853-885. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

7 Feb 2002

Figure 4 Anatomy of C-type lectin-like domains of NK receptors. Ribbon diagrams of Ly49A (PDB entry code 1QO3) (57), Ly49I (1JA3) (N Dimasi, MW Sawicki, LN Reineck, Y Li, K Natarajan, DH Margulies & RA Mariuzza, unpublished data), NKG2D (1HQ8) (56), CD69 (1FM5) (54), CD94 (1B6E) (53), and MBP-A (1BCH) [Weis, 1991 #1032]. The secondary structural elements are colored as follows: βstrands blue, α-helices red, and loop regions gold. The disulphide bonds are shown in green as ball-and-stick representation. The Ca2+ ions bound to MBP-A are drawn as magenta spheres.

Annu. Rev. Immunol. 2002.20:853-885. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

7 Feb 2002

9:12

AR

AR152-26-COLOR.tex

AR152-26-COLOR.SGM

LaTeX2e(2002/01/18)

P1: GDL

Figure 5 Structures of the Ly49A, Ly49I and NKG2D dimers and regions of interaction with MHC-I. Side views of the Ly49A (A), Ly49I (D) and NKG2D (G) homodimers with N-termini at the bottom. In these ribbon models, the β-strands are green and the α-helices are red and yellow. Top views of the Ly49A (B), Ly49I (E ) and NKG2D (H ) dimers. In these surface representations, the regions corresponding to the binding sites for MHC-I on Ly49A and Ly49I, or to the MICA binding site on NKG2D, are outlined in magenta. For Ly49A, the surface on each monomer at the Site 2 interface with H-2Dd (57) is shown. The corresponding regions of Ly49I, based on the Site 2 interaction in the Ly49A/H-2Dd complex, are hypothetical since no Ly49I/MHC-I complex structure is available. Side views of the Ly49A (C ), Ly49I (F ) and NKG2D (I ) dimers in which the distances between corresponding features defining the binding surfaces of these NK receptors (58) are indicated.

Annu. Rev. Immunol. 2002.20:853-885. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

7 Feb 2002

9:12

AR

AR152-26-COLOR.tex

AR152-26-COLOR.SGM

LaTeX2e(2002/01/18)

P1: GDL

Figure 6 Structure of Ly49A in complex with its MHC-I ligand H-2Dd (PDB entry code 1QO3) (57) illustrating the two sites of interaction. Site 1 is indicated as a circle and site 2 as a square. The two Ly49A subunits are shown in cyan and light blue. The H-2Dd heavy chain is gold, the peptide is blue, and β2m is gray. (An animation of the Ly49A/H-2Dd complex structure is offered as supplementary material in the online version of this chapter or at http://www.annualreviews.org/.)

Annu. Rev. Immunol. 2002.20:853-885. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

7 Feb 2002

9:12

AR

AR152-26-COLOR.tex

AR152-26-COLOR.SGM

LaTeX2e(2002/01/18)

P1: GDL

Figure 7 Interaction of NKG2D with the MHC-I homologs MICA, RAE-1 and ULBP. (A) Ribbon representation of the human NKG2D/MICA complex (PDB entry code 1HYR) (58). (B) The mouse NKG2D/RAE-1b complex (1JFM) (P Li, G McDermott & RK Strong, submitted). (C ) The mouse NKG2D/ULBP complex (S Radaev, B Rostro, AG Brooks, M Colonna, & PD Sun, submitted ). MICA, RAE-1β and ULBP are gold; the NKG2D monomers are blue and pink. (An animation of the NKG2D/MICA complex structure is offered as supplementary material in the online version of this chapter or at http://www.annualreviews.org/.)

Annu. Rev. Immunol. 2002.20:853-885. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

7 Feb 2002

9:12

AR

AR152-26-COLOR.tex

AR152-26-COLOR.SGM

LaTeX2e(2002/01/18)

P1: GDL

Figure 8 Recognition of MHC-I by the Ly49A, KIR2DL and NKG2D NK cell receptors. The KIR2DL2/HLA-Cw3 and NKG2D/MICA complexes (50, 58) were superimposed onto the Ly49A/H-2Dd complex (57) using equivalent Cα atoms of the α1 and α2 domains of the MHC-I molecules. For clarity, the only MHC-I molecule shown is H-2Dd. The H-2Dd heavy chain is gold, the peptide is blue, and β2m is gray. The Ly49A monomers interacting at Site 2 are cyan and light blue, KIR2DL is magenta, and the NKG2D monomers are green and pink. The overlapping area of KIR2DL and NKG2D is transparent.

P1: FRK

February 1, 2002

12:47

Annual Reviews

AR152-FM

Annual Review of Immunology Volume 20, 2002

Annu. Rev. Immunol. 2002.20:853-885. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

CONTENTS FRONTISPIECE, Charles A. Janeway, Jr. A TRIP THROUGH MY LIFE WITH AN IMMUNOLOGICAL THEME, Charles A. Janeway, Jr.

xiv 1

THE B7 FAMILY OF LIGANDS AND ITS RECEPTORS: NEW PATHWAYS FOR COSTIMULATION AND INHIBITION OF IMMUNE RESPONSES, Beatriz M. Carreno and Mary Collins

29

MAP KINASES IN THE IMMUNE RESPONSE, Chen Dong, Roger J. Davis, and Richard A. Flavell

PROSPECTS FOR VACCINE PROTECTION AGAINST HIV-1 INFECTION AND AIDS, Norman L. Letvin, Dan H. Barouch, and David C. Montefiori T CELL RESPONSE IN EXPERIMENTAL AUTOIMMUNE ENCEPHALOMYELITIS (EAE): ROLE OF SELF AND CROSS-REACTIVE ANTIGENS IN SHAPING, TUNING, AND REGULATING THE AUTOPATHOGENIC T CELL REPERTOIRE, Vijay K. Kuchroo, Ana C. Anderson, Hanspeter Waldner, Markus Munder, Estelle Bettelli, and Lindsay B. Nicholson

55 73

101

NEUROENDOCRINE REGULATION OF IMMUNITY, Jeanette I. Webster, Leonardo Tonelli, and Esther M. Sternberg

125

MOLECULAR MECHANISM OF CLASS SWITCH RECOMBINATION: LINKAGE WITH SOMATIC HYPERMUTATION, Tasuku Honjo, Kazuo Kinoshita, and Masamichi Muramatsu

165

INNATE IMMUNE RECOGNITION, Charles A. Janeway, Jr. and Ruslan Medzhitov

KIR: DIVERSE, RAPIDLY EVOLVING RECEPTORS OF INNATE AND ADAPTIVE IMMUNITY, Carlos Vilches and Peter Parham ORIGINS AND FUNCTIONS OF B-1 CELLS WITH NOTES ON THE ROLE OF CD5, Robert Berland and Henry H. Wortis E PROTEIN FUNCTION IN LYMPHOCYTE DEVELOPMENT, Melanie W. Quong, William J. Romanow, and Cornelis Murre

197 217 253 301

LYMPHOCYTE-MEDIATED CYTOTOXICITY, John H. Russell and Timothy J. Ley

SIGNAL TRANSDUCTION MEDIATED BY THE T CELL ANTIGEN x

323

P1: FRK

Annu. Rev. Immunol. 2002.20:853-885. Downloaded from arjournals.annualreviews.org by HINARI on 09/01/07. For personal use only.

February 1, 2002

12:47

Annual Reviews

AR152-FM

CONTENTS

xi

RECEPTOR: THE ROLE OF ADAPTER PROTEINS, Lawrence E. Samelson INTERACTION OF HEAT SHOCK PROTEINS WITH PEPTIDES AND ANTIGEN PRESENTING CELLS: CHAPERONING OF THE INNATE AND ADAPTIVE IMMUNE RESPONSES, Pramod Srivastava CHROMATIN STRUCTURE AND GENE REGULATION IN THE IMMUNE SYSTEM, Stephen T. Smale and Amanda G. Fisher PRODUCING NATURE’S GENE-CHIPS: THE GENERATION OF PEPTIDES FOR DISPLAY BY MHC CLASS I MOLECULES, Nilabh Shastri, Susan

371

Schwab, and Thomas Serwold

395 427

463

THE IMMUNOLOGY OF MUCOSAL MODELS OF INFLAMMATION, Warren Strober, Ivan J. Fuss, and Richard S. Blumberg T CELL MEMORY, Jonathan Sprent and Charles D. Surh

GENETIC DISSECTION OF IMMUNITY TO MYCOBACTERIA: THE HUMAN MODEL, Jean-Laurent Casanova and Laurent Abel ANTIGEN PRESENTATION AND T CELL STIMULATION BY DENDRITIC CELLS, Pierre Guermonprez, Jenny Valladeau, Laurence Zitvogel, Clotilde Th´ery, and Sebastian Amigorena

495 551 581

621

NEGATIVE REGULATION OF IMMUNORECEPTOR SIGNALING, Andr´e Veillette, Sylvain Latour, and Dominique Davidson

669

CPG MOTIFS IN BACTERIAL DNA AND THEIR IMMUNE EFFECTS, Arthur M. Krieg

PROTEIN KINASE Cθ

709 IN

T CELL ACTIVATION, Noah Isakov and Amnon

Altman

RANK-L AND RANK: T CELLS, BONE LOSS, AND MAMMALIAN EVOLUTION, Lars E. Theill, William J. Boyle, and Josef M. Penninger PHAGOCYTOSIS OF MICROBES: COMPLEXITY IN ACTION, David M. Underhill and Adrian Ozinsky

761 795 825

STRUCTURE AND FUNCTION OF NATURAL KILLER CELL RECEPTORS: MULTIPLE MOLECULAR SOLUTIONS TO SELF, NONSELF DISCRIMINATION, Kannan Natarajan, Nazzareno Dimasi, Jian Wang, Roy A. Mariuzza, and David H. Margulies

853

INDEXES Subject Index Cumulative Index of Contributing Authors, Volumes 1–20 Cumulative Index of Chapter Titles, Volumes 1–20

ERRATA An online log of corrections to Annual Review of Immunology chapters (if any, 1997 to the present) may be found at http://immunol.annualreviews.org/

887 915 925